(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "The varnishes of the Italian violin-makers of the sixteenth, seventeenth and eighteenth centuries, and their influence on tone;"

Google 



This is a digital copy of a book that was preserved for generations on Hbrary shelves before it was carefully scanned by Google as part of a project 

to make the world's books discoverable online. 

It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject 

to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books 

are our gateways to the past, representing a wealth of history, culture and knowledge that's often difficult to discover. 

Marks, notations and other maiginalia present in the original volume will appear in this file - a reminder of this book's long journey from the 

publisher to a library and finally to you. 

Usage guidelines 

Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the 
public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing this resource, we liave taken steps to 
prevent abuse by commercial parties, including placing technical restrictions on automated querying. 
We also ask that you: 

+ Make non-commercial use of the files We designed Google Book Search for use by individuals, and we request that you use these files for 
personal, non-commercial purposes. 

+ Refrain fivm automated querying Do not send automated queries of any sort to Google's system: If you are conducting research on machine 
translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the 
use of public domain materials for these purposes and may be able to help. 

+ Maintain attributionTht GoogXt "watermark" you see on each file is essential for informing people about this project and helping them find 
additional materials through Google Book Search. Please do not remove it. 

+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are doing is legal. Do not assume that just 
because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other 
countries. Whether a book is still in copyright varies from country to country, and we can't offer guidance on whether any specific use of 
any specific book is allowed. Please do not assume that a book's appearance in Google Book Search means it can be used in any manner 
anywhere in the world. Copyright infringement liabili^ can be quite severe. 

About Google Book Search 

Google's mission is to organize the world's information and to make it universally accessible and useful. Google Book Search helps readers 
discover the world's books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web 

at |http : //books . google . com/| 



FRIENDS OF THE FOGG 

3 




uL FINE ARTS LIBRARY 



Fine Arts Library 

given through the 
generosity of 

Mr. E. W. Forbes 




-V^^^^^PV^ 







i 

* 

i 



THE • VARNISHES • OF • THE 



ITALIAN • VIOLIN - MAKERS • OF • THE 



SIXTEENTH • SEVENTEENTH • AND • • • 



EIGHTEENTH • CENTURIES • AND 



• • • • 



THEIR • INFLUENCE • ON • TONE 



• • • • 



By ... . 

GEORGE FRY, F.L.S., F.C.S. 



London : 
STEVENS & SONS, Limited, 
119 & 120, Chancery Lane . . 

1904. 



ART MUSSUil or 

URVAP^ tRsrrr 



a<x --rvo. 1^55" 



/2 






^94 



( iii ) 



PREFACE. 



The casual observations which led to the 
interesting research which forms the subject of 
this Treatise came as a ** bolt from the blue." 

Experts who have had constant opportunities 
for studying the varnish on old Italian instru- 
ments have, without exception, accepted the 
theory that it is an oil-varnish coloured to suit 
individual taste, although in no single case have 
their own descriptions of its appearance been 
found to support this conclusion ; various reasons 
and excuses have been suggested why modem 
violin-makers have been unable to reproduce it. 

No scientist or chemist appears to have given 

much attention to the subject, for the simple 

reason that no means have been available for a 

complete investigation. There are, no doubt, 

many accomplished chemists who find recreation 

in the study of music, but it is not to be expected 

that one of these would sacrifice a favourite and 

valuable old instrument for the Scike of obtaining 

a few grammes of varnish on the chance of being 

able to obtain from it some satisfactory informa- 

a2 



iv Preface. 

tion as to its composition ; especially as he could 
have no reason to suspect the veracity of the 
dogma that it was a mere concoction. 

Direct investigation of the causes of the colour 
effects of sunrise, sunset, of the rainbow, &c. was 
equally impossible ; these natural and familiar 
phenomena were, for a long time, interpreted in 
accordance with the evidence of the senses, until 
scientific discovery of the composition, refraction 
and polarization of light afforded explanations 
which satisfied all the conditions. 

It is not surprising that the old varnish, which 
could be judged only from its appearance, its 
softness, elasticity, and its solubility in alcohol, 
should have been interpreted in a similar simple 
fashion. 

The theory offered to explain its composition 
was accepted, for want of a better one; it has 
now been tested for at least half- a -century, 
varnishes have been constantly produced in 
accordance with it; they have not been found 
to satisfy all the conditions : in fact, do they 
satisfy one of them ? The modem varnishes are 
not like the old ones in body (pS,te), in trans- 
parency, in colour, in refraction, the instruments 
which they clothe are not comparable with the 
old ones in tone. 



Preface. v 

If from the apologists for the old theory 
information is sought as to the actual con- 
stituents of the varnish, the result is equally 
unsatisfactory: a gum for the basis is vaguely 
suggested, btit eludes definition ; colouring 
matters, soluble in alcohol, are mentioned, but 
the difficulty is overlooked that alcohol, of suffi- 
cient strength for their solution and admixture, 
was certainly not available in the sixteenth 
century ; the further suggestion, that red colour- 
ing matters, soluble in oil of turpentine, existed 
then, but have ceased to exist now, will not peiss : 
for it is easy to show that such red colouring 
matter is a scientific impossibility. 

From every point of view, therefore, the 
dogma which has been so long and so persis- 
tently promulgated is found to be unsatisfactory. 

The explanation of the mystery which is now 
offered is that the old violin-makers used as the 
constituents of their varnishes the natural pro- 
ducts of trees (conifers) and plants (flax) growing 
in their immediate vicinity, abundant and easily 
procured ; that they were simple varnishes com- 
posed of resin and turpentine, or of these two 
substances and linseed-oil; that the various 
apparent colours were due to optical effects 
naturally arising from variations in the details 



vi Preface. 

of the preparation of the varnishes; that the 
differences in their physical qualities arose from 
the same causes. 

On careful examination it will certainly be 
found that this new theory satisfies every one of 
the required conditions and fully explains all the 
observed facts. 

All that I ask from my readers is an impartial 
judgment on the evidence; magna est veritcis, 
et praevalebit ! 

It is conceived that my own opinion is now 
the less likely to be warped from the considera- 
tion that, although my researches originated in 
the varnish question, this has long since become 
subsidiary to other and more absorbing problems 
which have arisen, which have completely over- 
shadowed the original one. From a scientific 
point of view (physiological and chemical), the 
constitution of the terpenes, whence they come 
and whither they go, are questions which will be 
fascinating long after the minor problem, of the 
constitution of the old varnishes and their 
influence on tone, has been set at rest. 

G. F. 

Berwick-on-Tweed, 1904. 



( vii ) 



CONTENTS. 



CHAPTER I. 

Introduction. 

Evolution of violins from viols — Introduction of the 
sound-post — Probable influence of the Church on 
the evolution of violin and varnish — ^The varnishes 
were oil-varnishes, the early ones perhaps excepted — 
Brown colour of early instruments and its recru- 
descence — Improvement of colour attends that of 
form — ^Extinction of the varnish about 1750 — Its 
quality as excellent as its colour — Differentiation of 
Cremonese and Venetian varnish — Introduction of 
Neapolitan varnish — ^Assumption of composition of 
the varnish, derived from superficial examination, 
generally accepted — Improvement suggested by 
Eugene Mailand — Defects in imitations supposed 
to be remedial by age — Observations which caused 
suspicion of the correctness of established dogma — 
Definition of dichroism — Origin and progress of a 
research on a new hypothesis — Scope and arrange- 
ment of the present treatise pages i — 8 



CHAPTER II. 
Descriptions of the old varnishes, modern opinions 

AND imitations. 

Correct description by J. B. Cartier, violinist (b. 1765) 
— Description by de Try — Opinion and descrip- 
tions by George Hart — Descriptions by F^tis 
— Opinions of Antoine Vidal under the influ- 
ence of J. ?. Vuillaume — Influence of Mailand on 



viii Contents, 



J. B. Vuillamne — Pre-eminence of Stradivari ; retro- 
gression of his pupils — Mailand's treatise — his un- 
founded preconceptions — Quotations from authors 
of the 1 6th, 17th, and i8th centuries — ^their irrele- 
vance to his contentions — Mailand's conception of 
a Cremonese varnish — Useful evidence adduced by 
Mailand — early use of water-bath and boiled linseed- 
oil — effect of heat in causing viscosity — pine- and 
fir-products used by old varnish-makers — ^ignorance 
of methods of colouring — ^The question of a secret 
and the possibility of its maintenance — Mailand con- 
cludes that the colours used were resins or resinous — 
is ignorant of their mode of introduction — Colouring 
matter admitted to be insoluble in oil of turpentine 
and other essences — Observation by Mailand that 
thin films of old varnishes are yellow ; he attributes 
this to preliminary sizing — His reasons for sizing 
controverted — Descriptions by Messrs. Hill — Rough 
diagram showing the movements of the various char- 
acteristics of the varnishes — ^AU the descriptions are 
shown to be referable to refraction eflfects, not to 
coloured varnish — Similarity between the dichroism 
of sunrise and sunset and that of the varnish ex- 
plained — Unappreciated service of Mailand. .pages 9 — 31 



CHAPTER HI. 

The tone of violins. 

Varnish affects tone — ^Tone of unvarnished violins transi- 
tory ; confirmation of this view by Vidal (Vuillaume), 
Sibire (N. Lupot), Savart — ^Description of hypo- 
thetical vibratory diaphragms of varying properties 
— Old varnish penetrated the wood — ^E. J. Payne's 
description of the penetrating quality of Stradivari's 
varnish — ^The views of F61ix Savart quoted and dis- 
cussed ; climatic influences ; relation of wood to 



Contents. ix 



water and varnish ; inconstancy of weight of wood 
due to moisture — Long digression to illustrate the 
probable effects of continued vibration on the mutual 
adhesion and aggregation of the elongated cells of 
wood — deterioration of tone attributed to dislocation 
of the cells — ^such dislocation prevented by satura- 
tion with varnish — Decline of response with constant 
playing ; its causes — ^Disintegration and resumption 
of cohesion of resins — Complex condition of the 
diaphragms of varnished violins — ^Effect of rest — 
Properties of varnish affect quality of tone — ^Effects 
of age — Conditions which affect amplitude of vibra- 
tion also influence quality of tone pages 32 — 52 



CHAPTER IV. 

On the manufacture of oil-varnish. 

Definition of perfect and imperfect varnishes — Skill of 
experts engaged in manufacture of oil- varnish — 
Laurent Naudin's treatise — His descriptions of the 
processes — Different qualities of copal ; their influ- 
ence on the products — Want of conduction of heat 
by resins; its influence on the determination of 
fusing-points — Fusion of hard copal, solution in 
linseed-oil, dilution with oil of turpentine — Critical 
condition of the finished varnish — Failure of attempts 
to produce varnish by more rational methods — Infor- 
mation given by Naudin as to the chemistry of 
copal and its derivatives — ^The bearing of this infor- 
mation on the preparation of varnish — Fusion a 
misnomer for the splitting of copal by heat — 
Suggested explanation of the observed phenomena 

pages SZ—do 



V. 



Contents. 



CHAPTER V. 

Oil-varnish from turpentine dbrivativbs and its 

application. 

Observation of dichr5ism in an old varnish — Similar 
dichroism observed in a panel of pitch-pine — Oxi- 
dized pitch-pine exhibits the various colours of the 
old varnishes — When oxidized alone this wood 
resembles in colour the varnishes of Cremona and 
Venice ; treated with linseed-oil previous to oxida- 
tion the resemblance is to those of Naples — ^These 
effects discerned to be due to oxides of turpentine 
and their derivatives — Resin contemporary with the 
violin — Geographical considerations allot pine pro- 
ducts to Brescia, larch products to Venice — ^Varnish 
from pine products called Cremonese, from larch 
Venetian — Linseed-oil selected as the drying oil — 
Choice of an oxidant — Reason for the selection of 
nitric acid — ^Preparation of turpentine varnish — 
Deductions from results — Its want of cohesion — 
Reasons for introduction of linseed-oil — Preparation 
of oil-varnish from colophony oxidized (dry way) — 
Fusing point of colophony — Oil varnish from colo- 
phony (wet way), Venice turpentine and oil of 
turpentine — Nitrification of* the latter described — 
Preparation of Neapolitan varnish — Preparation of 
deep-coloured varnishes by dehydration — Difficul- 
ties attending the use of the sand-bath ; introduc- 
tion of the water-bath and water-oven — ^Variety 
caused by variation of the different factors con- 
cerned in the manufacture of varnish — ^The propor- 
tion of boiled oil and its quality — Disadvantages of 
the Neapolitan process — Relative proportion of 
resin derived from colophony, Venice turpentine 
and oil of turpentine — ^Appreciable difference in 
products from these three substances — Pale var- 
nishes easy of application, dark ones difficult ; signi- 
ficance of resemblance in this respect of old and 



Contents. xi 



modem — ^Explanation of difficulties besetting the 
preparation of limpid dark varnish — Influence of 
absorption by wood on even application — Best 
method of rubbing down — Influence of wood on 
colour effects . . , pages 61^-92 



CHAPTER VI. 

Interesting chemical problems which have presented 
themselves during the progress of this research. 

Dichr5ism, its earliest observation in terpene products 
by Berthelot — Its restricted meaning — Production 
of a dichroic solution from alcohol-pinene by the 
action of nitric acid at ordinary temperature — Solu- 
bility of colophony in aqueous alcohol — Division of 
alcoholic solution of colophony into two parts by 
weak nitric acid : one of which, the most soluble, is 
dichroic ; the other is not, but becomes slightly so 
by further oxidation — Production of oxy-pinene, a 
dichroic resin dissolved in a volatile oil, optically 
inactive — ^The phenomena attending the production 
of a laevo derivative from dextro-pinene — Theory of 
the formation of resins by Prof. W. A. Miller — Sug- 
gested modifications — Drying and solidification of 
varnish differentiated — Observations on the drying 
of four varnishes — Loss of weight followed by in- 
crease supposed to imply absorption of oxygen — 
Difference in solubility of oxy-terpenes and their 
dehydrated derivatives — Precautions required in 
dilution of varnish — Paradox arising from vis- 
cosity of a varnish requiring more dilution while its 
constitution demands less — Suggested explanation — 
Oil-varnishes probably often supersaturated solutions 
— Critical condition of a finished varnish attributed 
to change in constitution of the oil of turpentine 
used for dilution — ^The chemistry of the terpenes — 
Plea for further investigation, with reasons .. pages 93 — 1 1 1 



xii Contents. 



CHAPTER VII. 

Conclusion. 

Greneral conception of the old varnishes derived from 
descriptions by the best authorities — ^The influence 
of Stradivari on their improvement; the aim and 
purpose of his life — ^The course of evolution deduced 
from existing records — ^An explanation given of the 
various stages from their origin in Brescia down to 
the great epoch of Stradivari and its close in 1725 — 
Detailed account of the evolution of the modem var- 
nishes in my hands— My instruments compared with 
the representations of some of those of Stradivari — 
Striking similarity in shades of colour — Comparison 
of dates — Classification of Stradivari's varnishes into 
Cremonese, hybrid Cremonese-Venetian, and Vene- 
tian — ^Differences in tone due to the use of these 
different varieties — ^The object Stradivari had in view 
for changing his varnish — ^The evolution of decline — 
Reappearance of brown colour — Reasons for retro- 
grade movement — Final evolution of decline, the 
introduction of Neapolitan methods, its object and 
the difficulties encountered — Idios3mcrasies of the 
old and modem varnish compared — ^The effects of 
cases; their disadvantages compared with the old 
plzin of hanging by the scroll — ^A probable cause of 
sudden extinction — ^The question of the existence of 
a secret considered — ^A parallel case of a trade 
secret — ^Were the old vamishes unfit for other pur- 
poses as has been asserted ? — practical demonstration 
to the contrary — Appreciation of the loss sustained 
by the extinction of the old methods — ^The rush for 
old instraments and its deplorable results — ^Experi- 
ments on the tone of old violins versus new- ones — 
Remarkable results and the inferences to be drawn 
from them — ^The hope of the approach of a new era. 

pages 112 — 163; 

BIBLIOGRAPHY page 164, 



APPENDIX ; page 165 



THE 



jarmsltw of l\t Italian ^jiolin-mate^ 




Chapter I. 

INTRODUCTION. 

'HE gradual evolution of Violins (this term is 
used generically to include instruments of all 
sizes) from the ancient viols, traced after diligent 
research, has been well described by several writers. 
By means of workmen, unlearned but intelligent 
and skilful, this evolution appears to have been, 
in the main, a natural process from its origin 
in Brescia in the fifteenth to its perfection in 
Cremona in the beginning of the eighteenth 
century. 

The introduction of the sound-post, however, an 
innovation of the utmost importance, appears to 
stand apart from this natural evolution of form. 
M. Vidal refers this invention to the "Trompetta 
marina " (illustrations of this curious instrument are 
given by him and by Naumann). In this instrument 
one foot of the bridge rests firmly upon the table 
(sounding-board), while the other foot approaches 
it so closely as to facilitate the communication of 
vibrations ; the player uses the bow with one hand 

V. B 



2 Varnishes of the Italian Vtolin'tnakers. 

at a point not very far from the nut, while the notes 
are formed by the fingers of the other hand placed 
on the strings between the bow and the bridge. The 
notes thus produced must therefore have been all 
harmonics of the prime note of the string. 

Judging by the information which is available 
concerning the education of the violin-makers of 
Brescia, it is impossible to avoid the suspicion that 
the design of the "Trompetta marina" came from 
some more erudite source; the same may be said 
of the sound-post, whether it was derived from the 
above-mentioned instrument or not. 

In this, as in the subject about to be considered, 
the great influence of the Church of Rome, through 
members of her monastic institutions who were 
students of the arts and sciences, cannot be ignored. 

As far as it is possible to judg^ from the instru- 
ments which are still in existence, the evolution of the 
varnish in which the violins were clothed appears to 
have followed with equal steps that of form and con- 
struction. 

From existing descriptions, and from coloured 
illustrations, it is comparatively easy to form a con- 
ception of the varnish in its different stages without 
any very extensive acquaintance with the instruments 
themselves. Its general characteristics were mani- 
festly persistent. 

There is a consensus of opinion that the varnish is 
that commonly known as an " oil-varnish." To this 
opinion, however, there is a notable exception : M. F6tis 
describes the varnish of J. P. Maggini (working about 
1590 — 1640) as a "spirit-varnish" — "la plupart de 
" ces instruments sont vemis k Tesprit-de-vin, d'une 



Introduction. 



" belle couleur brun clair. Ce vemis est remarquable 
" par sa finesse." * (Vol. V. 400.) A reason for this 
divergence will be hereafter suggested. 

The earliest instruments of Brescia and Cremona 
were of a brown colour; although the later examples 
of Maggini are described and represented as inclining 
more towards yellow and orange, a brown tint always 
prevails. Among the later instruments of Cremona, 
Rome and other Italian cities, a recrudescence of this 
brown colour or tint is often to be observed. 

As the form and workmanship of the violins im- 
proved, the varnish gradually exhibited purer tints of 
yellow and orange, brown shades became more rare, 
until the warm tones of Stradivari and his contempo- 
raries appeared towards the end of the seventeenth 
century. 

From this time until the quality of the varnish 
began to decline, these tints of orange and red pre- 
vailed among the Cremonese and Venetian makers. 

About the year 1750, the varnish which the violin- 
makers had used for more than two centuries almost 
entirely disappeared, leaving no trace of its mode of 
manufacture. A few rare examples appeared a little 
later (before the close of the eighteenth century) ; these 
are supposed to have been produced at the solicitation 
of noble clients of the violin-maker (Guadagnini). 

Apart fi"om colour, the quality of this unrivalled 
varnish, the composition of which is now unknown, is 
excellent in every respect. When dry it still remains 
slightly soft, and is always elastic. If the wood on 

* The greater part of these instmments are varnished with a spirit 
varnish, of a fine light-brown colour. This varnish is remarkable for its 
fine quality. \ , 

b2 



4 Varnishes of the Italian Vtolin-makers. 

which it was spread was not perfectly seasoned at 
the time of its application, when the grain rose the 

varnish neither cracked nor fissured — it followed the ^ 

i 

most minute undulations of the wood. I 

I 

For the sake of brevity, the varnishes of Brescia 
and Cremona will be spoken of as Cremonese. The 
varnish of Venice is described as very similar to that i 

of Cremona. It is not certain that experts are capable 
of distinguishing the one from the other ; it will be I 

designated as Venetian. 

It has been observed that some of the varnish of 
Stradivari is a little harder (more inclined to chip) 
than the generality of the Cremonese (Hill's " Stradi- 
vari," 174). It would have been interesting had the 
information on this point been more definite, espe- 
cially if the dates and colour of the instruments 
possessing this peculiarity had been noted. 

At some undefined period after the varnishes 
already mentioned had reached their zenith, varnish 
of a slightly different character was introduced by the 
violin-makers of Naples, Bologna, &c. It is less 
transparent than the earlier varieties when of a red 
colour, and, generally speaking, is not so soft or 
elastic. Much of it is yellow in colour, some of it 
red, of a deeper tint than the warmest tones of the 
Cremonese or Venetian varieties. This varnish will 
be called Neapolitan to distinguish it. 

For a century the modes of preparation of these 
varnishes — Cremonese, Venetian, Neapolitan — have 
been lost. Violin-makers, chemists, amateurs, all 
sorts and conditions of men, have in vain attempted 
to re-discover the secret. 

Apparently firom superficial observation, an opinion 



Introdtution. 



was soon formed among experts that the old varnishes 
were coloured, that is to say, that they all consisted 
of a similar or common basis tinted by the introduction 
of adventitious, substantive colouring matters, ac- 
cording to the taste of the different makers. This 
natural deduction from appearances has been reiterated 
with so much confidence, by persons having an exten- 
sive practical knowledge of the old instruments, that 
it has hitherto been generally accepted. 

To produce successful imitations, however, in ac- 
cordance with this view of their constituents, was by 
no means easy, owing to the difficulties attending 
the admixture of colouring matters with oil-varnish : 
organic pigments being practically insoluble in oil of 
turpentine and in linseed-oil. 

These difficulties were in a great measure sur- 
mounted by means discovered by M. Eugene Mailand, 
an educated amateur, who published a book on the 
subject in Paris in 1859. There can be little doubt 
that this treatise is the foundation of all the most 
successful methods which have been used in modem 
times for the imitation of the old varnishes, although 
its influence has not, at all times, been acknowledged. 

It has been found that these imitations do not 
possess the peculiar optical properties of the original 
substance, neither have the instruments covered with 
them qualities of tone similar to those of the old ones. 
It is supposed that these acknowledged defects are to 
be attributed to the want of the beneficial effects of 
time and use on the wood of which the violins are 
made as well as on the varnish in which they are 
clothed. 

Although for a number of years, several excellent 



Varnishes of the Italian Violin-makers. 



old Italian instruments had been in my possession, it 
must be confessed that the varnish question did not 
attract my attention. For more than forty years I 
accepted, like the rest of the world, without question 
or demur, the dogma which emanated from experts in 
whose knowledge and judgment I placed unlimited 
confidence. This apathy may be excused by the con- 
sideration that a coloured varnish, a mere concoction, 
is totally devoid of scientific interest. 

Suspicion of the general soundness of the estab- 
lished dogma was, quite unexpectedly, aroused by the 
examination (in May, 1900) of a red varnish of the 
Neapolitan variety, which was discerned to be dis- 
tinctly dichroic. 

The significance of this observation was not imme- 
diately apparent, but, after some consideration of the 
matter, means were taken to obtain access to a number 
of Stradivari's instruments, of undoubted authenticity. 
The varnish of these was found to exhibit the same 
dichroic properties — red tints were evidently optical 
effects, the apparent colour was in no case due to 
substantive colouring matter or pigment. 

The scientific term " dichroism " appears to have 
been primarily used " to denote the property exhibited 
" by many double-refracting crystals, of exhibiting 
" different colours when viewed in different directions" 
(Watts' Die. of Chemistry, 1 869, Vol. II. 320) ; Berthelot 
used it in 1 853, 1 854, to denote some such (unexplained) 
property in some terpene derivatives (Ann. de Chim. 
Vols. 38, 39, 40). It is here used to denote that the 
colour of transmission (yellow) differs from that of 
refraction (red, rarely purple). 

The problem now presented was totally different 



Introduction. 



from that of the skilful concoction of a coloured varnish ; 
a research was instituted with a view to its solution. 
The results so far obtained are deemed of sufficient 
scientific interest to be worthy of publication. 

The direction of the experiments was influenced 
by close observation of terpene derivatives to be found 
in different kinds of wood and in other arboreal 
products, many of which exhibited dichroism, due 
consideration being given to the physiology of the 
various organs of trees and the probable chemical 
action of the cells of which they are composed. 
The processes employed were intended to be imita- 
tions of those of Nature. 

The extent of progress in the desired direction 
could only be estimated by the somewhat slow and 
laborious process of varnishing slips and panels of 
wood; eventually, for the same purpose, violins, 
violas, and violoncellos, of the best quality of wood 
and workmanship, were covered with the varnishes. 

As was expected, after many months of study and 
experiment, it was found that every variety of the old 
varnishes could be reproduced with facility from tur- 
pentine and linseed-oil without the admixture of 
colouring matter in any form. The conclusion arrived 
at is that the old types were pure oil-varnishes, consist- 
ing of terpene oxides (more or less dehydrated) dis- 
solved in linseed-oil (oxidized to varying degrees), 
finally diluted with oil of turpentine ; in the oldest 
forms the linseed-oil was probably omitted. 

In order that the question may be understood in 
all its bearings, it is necessary that a large amount of 
information should be given. By lay readers this 
must be studied with some attention, if they desire to 



8 Varnishes of the Italian Violin-^makers. 

be in a position to form a judgment on the evi- 
dence. 

In the first place, descriptions of the varnishes will 
be gathered fi'om many sources ; arguments in favour 
of the theory of coloured varnishes, especially those, 
of M. Mailand, will be quoted and criticised. 

Secondly, an attempt will be made to elucidate 
the interesting question of the influence of varnish 
on tone. 

Thirdly, technical information relating to the 
preparation of oil-varnish, derived fi^om the most 
competent authorities, will be presented. 

Fourthly, the preparation of dichroic oil-varnishes 
will be described. 

Fifthly, a chapter will be devoted to the con- 
sideration of a few points relating to the chemistry 
of varnish and its constituents which have arisen 
during the progress of the experiments. 

Lastly, the whole case will be simimed up, and 
the evolution of dichroic varnishes, in my hands, will 
be compared with that of the olden time. 

For convenience of reference, and for the benefit 
of readers who may desire to pursue the subject 
further or read for themselves the authors quoted, 
a list of books is given ; those more directly bearing 
on the questions involved are marked by an as- 
terisk (*). It is not suggested that this list approaches 
completeness. 



( 9 ) 



Chapter II. 

DESCRIPTIONS OF THE OLD VARNISHES — ^MODERN 
OPINIONS AND IMITATIONS. 




T is not a little surprising that, among the many- 
descriptions of the appearance of the old var- 
nishes, only one can be said to be perfectly correct ; 
this was given by a violinist, not by an expert. It is 
to be found in F6tis' book (Vol. I. 82). Writing of 
the violins made, by Andreas Amati (who was alive 
and working about 1550), M. F6tis writes : — " Cartier, 
"qui a vu deux de ces violons, affirme que rien ne 
"surpasse la perfection de leur travail. lis 6taient 
"revfitus d'un vemis k I'huile d'un ton dor6, avec des 
"reflets d'xm brun rouge4tre."* The Cartier here 
referred to was Jean Baptiste Cartier, bom at Avignon, 
May 28, 1765, who came to Paris in 1783, where he 
became a pupil of Viqtti, afterwards violinist to Marie- 
Antoinette. He died in Paris in 1841. The violins 
in question formed part of the Royal collection — 
"les vingt-quatre violons du Roi" — so frequently 
mentioned in history. 

This is an undoubted description of a dichroic 
varnish. 

For another description which approaches this for 

* Cartier, who saw two of these violins, affirms that nothing could 
surpass the perfection of their workmanship. They were clothed in an 
oil-yamish of a golden tone, with reflections of reddish brown. 



10 Varnishes of the Italian Violin-makers. 

correctness of observation, we are indebted to M. J. 
Gallay, editor of the new edition of the Abb6 Sibire's 
book. The observer is again not an expert, but a 
violoncellist. M. Gallay writes (179) with reference 
to certain violoncellos, made by Antonio Stradivari, 
which were then in Madrid (the translation is mine): — 
" One of these basses, according to the account which 
" has been given to us by a French artiste, M. de Try, 
"a distinguished violoncellist, who has had, during a 
"tolerably long residence in Madrid, the good fortune 
"to play on these instruments, seems to us to be truly 
"unique. It is a specimen of the great period (1725) ; 
"the varnish, of a fine bright red, comes out bril- 
liantly on a first coat of amber yellow" ("le vemis, 
d'un beau rouge vif, ressort brillamment sur une 
premise couche jaune ambr6 "). 
Here again is an unmistakable description of 
dichroism. 

The opinion and descriptions of our own country- 
man, Mr. George Hart, will next be given, taken 
firom his well-known book. 

At page 36 et seq,j Mr. Hart expresses the opinion 
that, "although quite separable in one particular, 
"'which is, the depth of their colouring, the Brescian, 
" Cremonese, and Venetian have, to all appearance, 
" a common basis." In examining the Brescian 
varnish he finds "an almost complete resemblance 
" between the material of Gaspard di Salo and that of 
his coadjutors, the colouring only being different." 
Upon turning to the Cremonese, we find that 
Joseph Guamerius, Stradiuarius, Carlo Bergonzi, 
" and a few others used varnish having the same 
" characteristics, but, again, different in shade." He 



u 
u 



it 
u 



Old Varnishes and Modem Imitations. 11 



u 
u 
u 



u 
a 
ti 
u 



describes "the Brescian as mostly of a rich brown 
" colour and soft texture, but not so clear as the 
Cremonese. The Cremonese is of various shades, 
the early instruments of the school being chiefly 
amber-coloured, afterwards deepening into light red 
of charming appearance, later still into a rich brown 
" of the Brescian type, though more transparent," 
The Venetian is also of various shades, chiefly light 
red, and exceedingly transparent. The Neapolitan 
varnish (a generic term including that of Milan and 
a few other places) is very clear, and chiefly yellow 
in colour, but wanting the dainty softness of the 
" Cremonese." 

Excluding the Neapolitan, Mr. Hart looks upon the 
famous varnishes as having a common basis, coloured 
according to the taste and skill of the individual 
makers. This is a very natural conception : it forms 
the preponderating idea which induces his theories as 
to the origin, evolution and disappearance of the 
varnishes. As to the notion that " the varnish was 
" composed of a particular gum quite common in 
" those days, extensively used for other purposes 
" besides the varnishing of violins," reference will 
hereafter be made to it when the same suggestion is 
expressed in more confident terms by M. Mailand. 

The following remarks on the varnishes of different 
makers are selected firom those which appear in 
Mr. Hart's book : — 

" Amati, Antonius and Hieron3rmus, sons of 
Andrew, 1570 — 1635. The. varnish on the earlier 
specimens is deeper in colour than that found on 
" the later ones, which have varnish of a beautiftil 
" orange tint, sparingly laid on, and throwing up 



it 



12 Varnishes of the Italian Vio tin-makers. 



u 



the markings of the wood with much distinct- 



" ness." 



it 



Amati, Nicholas, 1596 — 1684. The bellies are 
" of a soft, silken nature, and usually of even grain. 
" A few of them are of singular beauty, their grain 
" being of a mottled character, which, within its 
" transparent coat of varnish, flashes light here and 
" there with strange force." 

" Bergonzi, Carlo, 17 18 — 1755. It is sometimes 
" seen to be extremely thick, at other times but 
" sparingly laid on ; often of a deep, rich red colour, 
** sometimes of a pale red, and again, of rich amber, 
" so that the variation of colour to be met with in 
" Bergonzi's violins is considerable. We must con- 
clude that his method of varnishing was scarcely 
so painstaking as that of his fellow-workers, if we 
may judge from the clots here and there, particu- 
larly on the deep-coloured instruments ; but, never- 
theless, now that age has toned down the varnish, 
" the effect is good." 

" Guamerius, Andreas, Cremona, 1630 — 1695. 
" The vairnish is much varied, but is generally of a 
" light orange colour of beautiful hue ; it sometimes 
" has a considerable body, but when so, lacks the 
" transparency of light-coloured varnishes." 

" Guamerius, Petrus, Cremona and Mantua, 
" 1690 — 1728. The varnish is superb. Its quality 
" is of the richest description, and its transparency 
" unsurpassed. Its colour varies ; it is sometimes of 
" a golden tint, sometimes of a pale red, on which the 
" light plays with delightftil variety." 

" Guamerius, Joseph Anthony, better known as 
" Giuseppe del Jesu, Cremona, 1683 — 1745. In some 



« 
ti 
ti 

n 
ti 



Old Varnishes and Modem Imitations. 13 

" cases the lustre of the wood of the backs, set in its 
" chasing of deep amber, that unrivalled varnish, 
" may be likened to the effect produced by the setting 
** sun on cloud and wave." Mr. Hart here quotes 
Byron's description of sunset in "Childe Harold." 
" The varnish on such instruments is of a rich golden 
" hue, highly transparent ; it is lightly laid on." 

" Stradiuarius, Antonius, Cremona, 1644 — 1737. 
" The varnish on the instruments belonging to the 
"period under consideration (1686 — 1694) is very 
" varied. Sometimes it is of a rich golden colour, 
" deliciously soft and transparent ; in other instances 
" he has used varnish of a deeper hue, which might 
" be described as light red, the quality of which is 
" also very beautiful." ..." The splendour of 
" the wood is unsurpassed in any violin, ancient or 
" modern, and it was named the ^ Dolphin ' from the 
" richness and variety of the tints it gives to the 
" varnish." 

In spite of his theory of coloured varnishes, it is 
manifest that the colour effects which he describes are 
those of dichroism. In other respects, the truth and 
significance of Mr. Hart's excellent descriptions will 
be hereafter apparent. 

M. F6tis, writing of Antoine Stradivari, represents 
him, up to 1670, as working under the influence of 
Nicholas Amati ; from 1670 to 1690 as more occupied 
in experiments and meditations on the perfecting of 
his art than in making instruments for sale. In 
1690 came an epoch of transition — "His varnish is 
more coloured." From 1700 to 1725, the instruments 
which leave his hands are so many perfect works — 
" Le bois, choisi avec le discemement le plus fin. 



14 Varnishes of the Italian Violin-makers. 



" r^unit k la richesse des nuances toutes les condi- 
" tions de sonorit6. Les beaux tons chauds du vemis 
" de Stradivarius datent de cette 6poque : la pAte en 
" est fine et d'une g^ande souplesse.* From 1725 to 
" 1730 the varnish is more brown." 

Another writer who is deservedly much esteemed 
as an authority on the Violin is M. Antoine Vidal. 
He was a friend of the late M. J. B. Vuillaume, the 
well-known instrument maker of Paris, who, together 
with M. Greorges Chanot, saw most of the instruments 
which Louis Tarisio brought anniially fi-om Italy, fi^om 
about the year 1827 onwards. When Tarisio died, in 
1854, M. Vuillaume bought of his heirs the (about) 
250 violins, &c. which he had accumulated. Among 
these were some of the finest instruments in existence. 
The dealers and connoisseurs of Paris, at that time, 
must have enjoyed unusual opportunities for studying 
many fine instruments, in their original condition. 

Much of what M. Vidal writes about the varnish 
will be translated and abbreviated ; when his remarks 
are of considerable importance they will be quoted in 
his own words. He is of opinion that the varnish has 
certainly an influence on the tone. If, on the one 
hand, it ought to have sufficient solidity and durability 
for the purposes of preservation, it must, on the other 
hand, possess such elasticity as to allow the parts of 
the instrument which it covers perfect liberty of move- 
ment ; without that, it would act as a mute. 

He repeats the old story of the necessity for making 

* The wood, chosen with the finest discrimination, unites to richness 
of shades all the conditions of sonorousness. The beautiful wann tones of 
the varnish of Stradivarius date from this period. Its body is fine and of 
great elasticity. 



Old Varnishes and Modem Imitations. 15 

a basis varnish with gums of the proper degree of 
hardness, and then colouring it with sandal wood, 
dragon's blood, &c., &c. The colouring is of less 
importance, although demanding much care. "All 
" that offers great difficulties, because, in the trials 
" (experiments), the greater part of the substances 
" treated together often do not agree, and everjrthing 
" has to be begun over again." 

Spirit varnishes have been tried; it is said that 
Louis Gxiersan, pupil of Boquay, was the first French 
maker to use it. It had the great advantage of drying 
quickly, a few days sufficing for varnishing an instru- 
ment. But the varnish encased the instrument as if 
with a carapace — sonorousness was wanting. These 
varnishes had to be abandoned. 

M. Vidal speaks with great praise of the researches 
of M. Eugfene Mailand, which became public know- 
ledge in 1859. I^ ^ footnote he makes the following 
interesting statement: — "Our great ^luthier,' M. J. 
B. Vuillaume, has affirmed to us that this book of 
M. Mailand had completely modified his ideas on 
" varnish, and that fi-om that time he had made 
" important changes in his method. It is easy to 
" convince oneself of this by examining the instru- 
ments which left his hands during the last fifteen 
years of his life. The varnish is incontestably 
superior to all that he had made before." 
M. Vidal further writes : — " Among the old Italian 
" masters, A. Stradivari is the one who, firom this 
" point of view, as fi"om many others, has reached the 
" greatest cleverness and perfection ; his varnish is a 
" model both in respect to its colour as well as to its 
" fine and transparent body. That of Joseph Guameri 






u 
it 



u 

€1 



16 Varnishes of the Italian Violin-makers. 

" del Jesii, when it is examined attentively, seems to 
" be a little thicker ; but it differs above all in the 
" nuance (hue), which is deeper. We shall see further 
" on, that the pupils of Stradivari commenced, in this 
" respect, to wander from his principles." 

" Carlo Bergonzi, among others of whom we shall 
" soon have occasion to speak, employed a varnish 
" sensibly more heavy, and which has given to certain 
" of his instruments that crusted appearance, much 
appreciated by amateurs, but which, in reality, is 
the beginning of decline (commence la decadence). 
** New resins are already employed, and starting from 
" this epoch, that is to. say, towards 1750, the methods 
** of the great masters begin to be completely lost. 
" As we have said, the progress of the (varnish) 
" industry makes itself felt ; the better will kill the 
" good." 

M. Vidal afterwards speaks of the violoncellos of 
Carlo Bergonzi: "with their wood so admirably 
" chosen, their red-brown (rouge-brun) varnish a little 
" thick, crackelled, but nevertheless pleasing to the 
" eye." 

He speaks of the varnish of Jean Paul Magini as 
being " of a light brown yellow (brun jaune clair).'* 

After a very long search, a copy of the reprint 
(1874) of M. Mailand's book was at last secured. I 
quite concur with M. Vidal in his estimation of it. 
Although I find myself unable to agree with M. 
Mailand's reasoning and his conclusions, it is readily 
admitted that this work stands alone as an honest 
and praiseworthy attempt to solve the mystery sur- 
rounding the old varnishes ; an attempt which must 
have required a large amount of labour and patience. 



Old Varnishes and Modem Imitations. 17 

M. Mailand writes (9) : — "We have already said 
** that the art of making varnishes at the epoch when 
" the great masters lived was in a nascent state. 
" Nevertheless, we are certain that they employed 
" those which they found in commerce and which 
" they appropriated to their wants. It is this which 
" explains to us why these also have sometimes 
changed, when new formulae were produced. But 
if they have left one formula for another, they have 
" always remained under the conditions of elasticity, 
which they had recognized as necessary. As to the 
colouration, it has little effect on the quality of the 
" varnish as body (elle importe peu i la quality du 
" vemis comme p4te), if one may so express oneself; 
" and if they have often varied in this respect, that 
" could only depend on the demand of the buyer, on 
" their taste, or on the more or less solidity of the 
" colouring matters." 

M, Mailand gives no evidence whatever in support 
of his certainty that the old masters used varnishes 
which they found in commerce. I agree that the 
" conditions of elasticity " remained practically con- 
stant, for the simple reason that the constitution of the 
varnish was the same from start to finish. As to his 
astonishing statement that the colouring matter would 
have little effect on the quality of the varnish ( "comme 
p4te *'), I am sorry to differ from him. The very large 
proportion of pigment which is required to be mixed 
with a pale, soft resin to produce a deep-coloured 
varnish has a most important and deleterious effect 
on the quality of the body of the varnish. 

He repeats, on page 1 2, that he " is well convinced 

V, 









it 
it 



18 Varnishes of the Italian Violin-makers. 

" that they (the violin-makers) could not have em- 
" ployed other than the varnishes then known in 
" industry." " Then," he fixes as between the years 
1550 and 1740. 

He then gives receipts published by Alexis, the 
Piedmontese ("Secrets des Arts") in 1550; by Fiora- 
vanti in 1564; by Auda in 1663; by Zahn in 1685; 
by Morley (Christophei^ in 1692 ; by Coronelli in 1693 ; 
by Pomet in 1694; by Bonanni in 17 13. As to the 
latter, he writes : — " We have found there precious 
" information which will allow us to conclude with 
" certainty." (" Nous y avons trouv6 de pr6cieux 
renseignements qui nous permettront de conclure 
avec certitude"). 

To conclude what with certainty ? Presumably 
what he has already set forth — that the great masters 
employed varnishes which were well known, at least 
to be found, in commerce. As this statement has been 
made by other writers besides M. Mailand, it will be 
necessary to ascertain if he produces any evidence to 
substantiate it ; for assuredly if he does not, no other 
writer does. 

It is submitted that what he has to do is to prove 
that a varnish was known in commerce which re- 
sembled that of the earliest masters of Brescia, at the 
time that they lived (manifestly common knowledge 
in 17 13 would not be relevant to Gasparo da Sal6 and 
Maggini) ; further, that the violin-makers improved 
or changed their methods as the knowledge or experi- 
ence of the outside world progressed. A literal 
translation of his own words is repeated: — "They 
" employed those which they found in commerce, and 
" which they appropriated to their wants. It is this 



Old Varnishes and Modem Imitations* 19 



u 



which explains to us why these also have sometimes 
changed when new formulaB were produced." 
The formulae quoted from the several authors are of 
little interest in relation to the contention which has 
been so confidently put forward by M. Mailand, 
Before endeavouring to trace a Cremonese varnish 
among these very crude prescriptions, some definite 
conception must, of necessity, be formed of the con- 
stitution of such a varnish. To avoid any difficulty 
on this point M. Mailand shall be met on his own 
ground; here is his prescription for a Cremonese 
varnish (156) : — 

" Mastic in tears - - - - 10 grammes 
" Soft: Dammar ('Dammar friable') 5 grammes 
" Essence, coloured according to one 
of the formulae already given - 100 cub. cent." 

The coloured essence prescribed in this formula 
requires some explanation. Having fiixed on his 
resins and determined that they must be dissolved in 
oil of turpentine (not in alcohol), M. Mailand then 
proceeds to the question of colouration. He found 
that the colouring matters which he desired to incor- 
porate could not be dissolved in oil of turpentine 
(essence) . On the other hand, they were freely soluble 
in alcohol. But on mixing such an alcoholic solution 
with oil of turpentine, the two solvents refiised to 
combine. To get over this difficulty, M. Mailand 
exposed oil of txirpentine to air and light for six or 
eight weeks; it became oxygenated and resinous, 
much more soluble in alcohol than the original, newly- 
distilled oil of turpentine ; this he calls "fat essence" 
(" essence grasse"). In this fat turpentine or "essence 

c2 



20 Varnishes of the Italian yiolin-makers. 

grasse " he dissolved his mastic and dammar, and to 
this solution he added his alcoholic solution of gam- 
boge, sandal-wood, dragon's blood, &c. Having 
incorporated this mixture as well as possible, he 
proceeded to distill off or evaporate a good deal of 
the alcohol (he appears to have imagined that he 
eliminated the whole of it, in which he was mistaken). 

It is quite safe to assert, beyond all chance of con- 
tention, that there is not the least indication of the 
preparation of " essence gfrasse," of any such concoc- 
tion or process, in the several writers quoted ; neither 
in Alexis, Fioravanti, Auda, Zahn, Morley, Coronelli, 
Pomet, or Bonanni. Whatever credit is due for the 
ingenuity in producing this varnish, is due to M. 
Mailand. To assert, on the evidence presented, that 
this varnish was to be fou^d in commerce at or before 
the time of Bonanni (i 7 13), is to asseverate that which 
cannot possibly be sustained. 

M. Mailand came to his task with a settled con- 
viction that the violin-makers used a varnish which 
was an article of commerce in their time. How he 
acquired this idea there is no evidence to shew. But 
his practice exhibited a good deal of common sense ; 
he ascertained, from contemporary writers on the 
subject of varnish, what materials the violin-makers 
had at their disposal ; applying modem methods to 
the concoction of these, he attempted to produce 
varnish similar to theirs. 

If the authors whose works he perused knew the 
varnish which the violin-makers used, it might have 
been expected that they would have g^ven a receipt 
for this special purpose, as they did in other cases : 

?.^., Alexis No. 2, "To give fine lustre to paintings " ; 



Old Varnishes and Modem Imitations. 21 

No. 3, "To make a liquid to be used to varnish 
" paintings " ; Fioravanti No. 8, " This varnish is ap- 
^* plied to skins"; No. lo, "Turkish varnish, which 
" is also used to varnish sheet-iron and for printers." 
Why is there no mention of "varnish for violin- 
" makers" or "for the makers of musical instruments"? 
Why does not M. Mailand point out the particular 
prescriptions on which he relies for the support of his 
theory ? 

It very often happens that the evidence sought is 
not the evidence obtained ; so in this case, although 
M. Mailand did not get the evidence which he fought, 
he is good enough to g^ve me some evidence which I 
wanted : — 

The majority of the authors were members of the 
Church of Rome — " les Pferes " Auda, Zahn, Morley, 
Coronelli, " le R. P. (r6v6rend p^re) " Bonanni, 
J6suite. 

Alexis (Pi^montais), writing in 1550, in his No. 7 
says, " and cook the whole for the space of one hour 
** in the water-bath (au bain-marie) in a vessel of 
" glass well corked." 

This certifies that the water-bath and its use were 
known as early as 1550. 

Fioravanti (Bologna, 1564) writes of linseed -oi (for 
its conversion to a boiled oil) : — "The most common 
" way is to boil the oil until it bums the vanes of a 
feather dipped into it ; some add a crust of bread, 
according to the quantity of oil which they are 
treating, because it absorbs its grease and makes it 
more easy to dry." 

Boiled linseed-oil was evidently known before 

1564. 






22 Varnishes of the Italian Violin-makers. 

Fioravanti, in the formula quoted as No. 1 1, 

Linseed-oil i part 

Greek pitch {i.e.y resin from the fir- 

trees of Calabria) . . . • 2 parts 

Pine resin ipart 

writes, " boil until the varnish is viscous." 

It was therefore known in 1564 that continued 
heating causes resinous substances dissolved in 
linseed-oil to undergo change of viscosity, probably 
of colour. 

Father Zahn ("Oculus Artificialis," 1685) again 
prescribes the "bain-marie." 

From Alexis we further learn that, in 1550, he 
used, to make varnish, " pine resin fat and white " 
(? galipot), "turpentine of Venice," linseed-oil, as well 
as what he calls varnish of amber, which M. Mailand 
explains to be " a mixture of oil and of Greek pitch, 
"which is the produce of the fir-trees (sapins) of 
" Calabria." 

It will be observed, from all the prescriptions, 
how much the old varnish-makers used the resins 
from the pine and fir-trees, as well as Venice tur- 
pentine (from larch). 

The gums Alexis knew were benzoin, mastic, 
male incense, sandarac. For colouring, he mentions 
aloes, red sandal-wood, dragon's blood, madder 
steeped in tartar water, Campechy and Brazil wood, 
all dissolved in barber's ley, to which alum must 
be added, afterwards boiled down. Speaking of the 
" fastness " of colours, he adds : — " Linseed-oil kills 
"the semi-mineral colours, but not the mineral and 
"vegetable." The amount of his knowledge on 



Old Varnishes and Modem Imitations. 23 






colouring varnishes maybe estimated from this and 
from the fact that he prescribes cinnabar and orpi- 
ment (mercuric sulphide and arsenious sesquisul- 
phide). 

M. Mailand writes (71) : — "But, it may be said to 
" us, the Cremonese ' luthiers ' perhaps had the secret 
" of a varnish which was not in commerce and which 
" has been lost with them. To that we reply, that we 
well understand that a person may have been able, 
doing better than his competitors, to have a secret, 
" which he has carried away with him in dying ; but 
" that it is not supposable for an instant that a thing 
"which has existed for more than a century, and 
"which is transmitted during several generations of 
" luthiers, could remain secret. It must necessarily be 
" known to everybody, and has only been abandoned 
" at the moment when considerable progress is made 
" in the manufacture of varnishes. If it is objected 
" that there are differences in the varnish of such and 
" such a master, it is possible ; but that would only 
prove one thing, which is that some knew better 
how to appropriate and employ them than others. 
"These nuances make themselves felt, for the rest, 
" more in the colour than in the varnishes themselves, 
"which were all composed of tender resins. It is 
** easy to be convinced of this. It is only necessary to 
" rub violins of this epoch with spirit of wine ; there is 
" not one that will resist ; if this operation is repeated 
" on objects varnished with fat, hard copal varnishes 
" which are made to-day, they cannot be moved." 

It is agreed that the old varnishes were soluble 
to some extent in alcohol ; that they were not " fat, 
" hard copal varnishes." 






* 24 Varnishes of the Italian Violin-makers. 



ti 
u 
ii 



it 






it 
ii 

a 



As to the colours which they put into the varnish, 
they are so very transparent or translucent (tene- 
ment translucides) and allow the grain of the wood 
so well to be seen (laissent si bien lire la veine du 
bois), that we do not doubt that they employed 
"resins or resinous colouring matters (!! !). It is 
impossible for us to say by what means they 
managed to dissolve these in the essences" {i.e.f oil of 
turpentine or of aspic), "but the means is indifferent ; 
what matters, is to succeed, and we have managed to 
mix them so intimately in the oils, that they do not 
abandon them again after they have been incor- 
« porated." 

At page 91, M. Mailand writes : — " Essence of 
turpentine, it is known, does not combine or mix 
with any kind of colour, whether ground up in oil, 
pulverized or distempered in the varnish ; it is the 
" same with the essences of lavender, aspic, or rosemary, 
" — ^they are tinted a little more, but so feebly that they 
" do not leave any colouration when laid on." Con- 
sequently, he gets the colour into the varnish by the 
medium of alcohol ; it does not matter how he gets it 
in so long as he succeeds. 

M. Mailand observed what I also observed, 
although we draw different conclusions from our 
observations. If a varnish were coloured red, it 
would be always red, however thin the pellicle ; but 
he noticed that a thin pellicle was not red but yellow. 
He writes (85) : — " Preoccupied by the yellow tone 
" which is met with on many of the instruments 
" of the Italian luthiers, when the varnish is worn 
away, we have thought that they sized them with 
resinous gums soluble in alcohol, such as gamboge 



(C 

ii 



Old Varnishes and Modem Imitations. 26 

" and aloes, which, it is said, has the property of 
" preserving wood from becoming worm-eaten. These 
" resins dissolved in alcohol are an excellent size, 
" which cannot interfere with the liberty of the tables, 
" nor the desiccation of the wood, since the alcohol 
" evaporates at once." 

M. Mailand easily smooths over missing links in 
his evidence. Here is an interesting example (93) : — 
" Alexis does not speak of gamboge, which, however, 
" was known in his time." 

Having made up his mind to sizing as a way out 
of this diflBculty (the colour of the worn part being 
yellow, while the thicker pellicle appeared orange or 
red), M. Mailand proceeds to account for the usa of 
size as a preliminary coat — ^the instructed reader will 
observe how many mistakes he makes in endeavouring 
to support his theory. He writes (81) : — " Is it neces- 
** sary to size instruments before varnishing them ? 
" We do not hesitate to reply affirmatively ; for if they 
" were not sized the two or three first (sic !) coats of 
" varnish would enter the wood ; they would remain 
" completely dull until saturation of the wood, which, 
" not being able to absorb any more of it, would be 
" sized by the varnish itself : this is easy to explain. 
The wood, by its cellular canals, fulfils the office 
of a sponge ; it absorbs the most fluid part of the 
" varnish, and so much the more the longer it has 
" previously undergone desiccation ; in consequence 
" the instrument will be saturated with essence, with 
** a proportion of resins and of colouring matters, of 
** which the surplus will remain on the surface, at 
" least, if the varnish is very viscid. These resins 
" deprived of a portion of their essence, being left to 






26 Varnishes of the Italian Violin-makers. 



it 



it 

it 









themselves (se trouvant abandonn6es k elles-mfimes), 
" will dry promptly, losing the suppleness which it 
" ought to communicate to them. It would evidently 
" result from this way of doing, that more coats would 
be necessary to obtain the necessary brilliancy, that 
the varnish which we employ, into which no hard 
resins enter, would become very friable, since they 
would have lost, in part, the excipient" (curious 
word ! compare Fleming 270) "which ought to pro- 
tect them, and that to arrive at a suitable solidity, 
it would be necessary to give two or three coats 
" more ; that at length, when these last shall have 
become oxidised by time or worn by rubbing, those 
beneath will fall promptly into powder ; but the 
" most grave inconvenience would be, on the one 
" hand, of having caused the wood to absorb a certain 
" quantity of varnish, and on the other of having put 
" on to it a thickness more considerable than useftil ; 
" in consequence the sounds of the instrument would 
be dull, veiled, without brilliancy, and without carry- 
ing power ; if not for ever at least for a very long 
space of time. The woods employed in violin- 
making (lutherie) are not used until they are very 
dry, desiccation which can only be obtained after 
several years, from which it follows that in causing 
" the essence and the resins to penetrate the wood, 
" one would do a thing quite contrary to the end 
" which one proposed to oneself to attain." 

This singular and totally erroneous notion of what 
happens to wood in the process of drying or seasoning, 
M. Mailand puts quite clearly, at page 61*: — " For in 
" violin-making the wood which is the most dry is 
" esteemed ; that, in a word, which has lost its resin " 



u 
ii 
ii 
ii 
ii 



Old Varnishes and Modern Imitations. 27 

(car on recherche en lutherie les bois les plus sees ; 
ceux en un mot qui ont perdu leurs resines). 

It is needless to say that wood is apt to become 
more, rather than less, resinous in drying ; the tesins 
which are in the fresh wood are perfectly stable and 
always remain ; parts of the volatile essential oils are 
likely to become resinous by oxidation during the 
slow, drying process. 

M. Mailand's statement as to what happens to a 
varnish when spread on dry wood is pure imagination ; 
it is not so easy to dissociate the resins from the 
heavier parts of the essence as he supposes. 

M. Mailand is fully aware that his view — ^that it is 
necessary to prevent the wood of violins from absorb- 
ing varnish — is directly opposed to that of M. Savart. 

There is one singular circumstance connected with 
the experiments of M. Mailand to which attention 
must be directed. It will be remembered that he 
rubbed an old Italian violin with spirits of wine in 
order to ascertain if the varnish were soluble in this 

•s. 

menstruum or not. He does not say how he made 
this test, but it may be assumed that he moistened a 
piece of cotton or linen cloth with alcohol and applied 
it to the varnish. The varnish was soluble in alcohol ; 
consequently a small quantity of dissolved varnish 
was transferred to the surface of the cloth. If M. 
Mailand had observed the colour of this varnish on 
the cloth, he would have seen that it was not the same 
as the apparent colour of the thick pellicle, but exactly 
matched that of the thin film where the varnish was 
worn away, which he had observed to be yellow. It 
is evident that if he had made this simple observation 
all his theories would have been scattered to the winds. 



28 Varnishes of the Italian Violin-makers. 



u 



Either he neglected to make the observation, or, if he 
made it, he was so infatuated with his preconceived 
ideas that he simply ignored it. 

• «•«•' 

The description which the Messrs. Hill give of the 
Maggini varnishes on the Dumas instruments is as 
follows (G. P. Maggini, 63) : — "The varnish of the 
violin is fine and abundant, and most pleasing in 
colour, being of a peculiar golden yellow, subdued 
" in places with pale brown. It yields, when a good 
" light plays on it, effects delightful in their way, 
" embodying at once the fascinations of amber and of 
" gems such as the sardius and topaz, though no 
" doubt some part of this fascination is due to changes 
" effected by time." Of the viola :— " Time and men 
" have indeed dealt kindly with this noble instrument. 
Not only is the wood intact, but the varnish shows 
only slight sigfns of wear, and is of the finest quality 
" and most original colour, a rich golden brown, 
ever ready to flash under the magic of light into 
colour surprises which beggar words. In sunshine 
the back seems compact of * mjniads of topaz lights,' 
touched here and there with gleams of purple, 
though here again, no doubt, time has done its 
part " (66). 

This exhausts the various statements and opinions 
which it appears necessary to record ; I have read a 
great deal more, both interesting and amusing, but, 
for the present purpose, I have failed to induce myself 
to tctke seriously the views put forward {e.^^.y Haweis). 
From these descriptions and firom others to be 
found in the authors quoted, a rough diagram may be 
constructed which may in a general and imperfect 






ti 
u 
u 

€€ 



Old Varnishes and Modem Imitations^ 29 

way give a graphic idea of the movements of the 
different qualities of the varnish. 



Viscosity 



Transparenct/ 
Brown 



Bed .... 
Yellow . 
Brescia . 
Cremona 
Venice ., 
Naples . 




M5 O »0 

A.D. S? »« ^ 



Md O kO 



*0 *0 CO CO CO ' CO 






to 



It is admitted that the generally expressed opinion, 
that the different varieties of the varnish had a common 
basis, is well founded. That this basis was used for 
other purposes is also admitted ; but this is quite a 
different thing from admitting that the basis was a 
gum or that the special varnishes under consideration 
were known in commerce or were used for other 
purposes, which is explicitly denied. 

All the observers, inadvertently, describe effects 
of colour caused by refraction; the most striking 
description is that given by Mr. Hart when he likens 
the appearance of the back of a violin by Joseph 
Guamerius del Jesii to "the effect produced by the 
" setting summer sun on cloud and wave." " The 



■» , 



30 Vamishts of the Italian Vtolin-makers. 

** varnish on such instruments is of a rich golden hue, 
" highly transparent ; it is lightly laid on." This 
simile is absolutely correct ; the conditions in the two 
cases assimilate very closely. 

In the case of the sunset, the effects are produced 
by refraction and polarization, caused by the passage 
of the solar rays, at a particular angle, through the 
refracting medium of the atmosphere ; when cloud or 
precipitated moisture lies in the path of the polarized 
rays, double refraction and dichroic effects ensue (in 
the shadows of clouds and waves complementary 
colour may be often observed). The highly refracting 
varnish of the violin represents its atmosphere ; the 
grain of the maple wood, which reflects the incident 
rays from constantly varying planes, replaces the 
clouds. A violin is selected of the golden hue which 
Mr. Hart specifies; the back is held in such a position 
that the incident rays of light and a line drawn from 
the observer's eye to the centre of the back are at an 
approximate right angle with the curved surfaces of 
the varnish; this position represents mid-day; the 
colour of the back is yellow interspersed with clouds of 
neutral colour. The angle which the longer axis of 
the back forms with the eye and the direction of 
the light is gradually made more obtuse; sunset 
approaches, warm tones slowly shew themselves in 
the clouds until, finally, the back of the violin appears 
to be composed of strata of red, red-brown, and 
yellow. 

How any educated person can conftise effects such 
as these with monochrome colour (even with two 
monochromes drawn one over the other) passes com- 
prehension. 



Old Varnishes and Modem Imitations. 31 

Mr. Hill's description of refraction effects is almost 
equally good, but he seems to be afraid to let himself 
go — ^he delineates forcibly with one hand while with 
the other he holds fast to a cherished theory. 

The excellent coloured representations of a few of 
the ancient instruments, for which the world is indebted 
to the Messrs. Hill, are equally fatal to the monochrome 
theory — ^the artist portrays the different thicknesses 
of the pellicle of varnish, the variety caused by light 
and shadow, not as shades of the same colour but as 
distinct colours. 

As with the sunset and sunrise, so with the varnish. 
Herein lies the charm — ^that the tints are constantly 
changing with varying conditions of intensity, quality 
and direction of light. It is absolutely impossible to 
imitate these effects by coloured imitations ; the 
notion that age will impart to these concoctions the 
properties of dichroic varnishes is quite illusory. 

M. Mailand did one great service to M. J. B. 
Vuillaume and his other followers, in that he taught 
them to use oxygenated turpentine in forming their 
mixtures; curiously enough, as far as can be dis- 
covered, neither he nor they appreciated the real 
significance of this proceeding. 



32 Varnishes of the Italian Violin-makers. 




Chapter III. 

THE TONE OF VIOLINS. 

'HE question whether the quality of the varnish 
applied to the wood of which violins (generic) 
are constructed has any decided influence on their 
tone, must be, without any hesitation, answered in the 
affirmative. 

The wood of which these instruments are made is 
of a light and absorbent character : it consists largely 
of sclerogen, without much encrusting or agglutinating 
matter. The wood of the belly, or table, which 
receives the primary impulse of vibration, through 
the medium of the bridge, is of the very lightest 
. description that can be procured ; at the same time, 
it is selected of regular and even grain. 

Apart from the question of appesurance and pro- 
tection, to some extent, from climatic and other 
deteriorating influences, it is generally admitted that 
an unvarnished violin will not permanently possess 
the necessary qualities for the production of good 
tone. 

M. Antoine Vidal (whose opinion, it may be sur- 
mised, was based, to a great extent, on the experience 
of his friend, M. Vuillaume) writes thus (the transla- 
tion is mine) : — " Experience replies clearly to these 
" questions : the violin finished and not varnished 
" has more power and mellowness in its tone ; but if 



The Tone of Violins. 33 






it remains in this virgin state, it becomes modified 
little by little, and, after a somewhat short time, 
" the tone becomes poor and feeble. It must 
" therefore be concluded that the varnish, while 
" giving a more pleasing appearance, conserves, pre- 
" serves, and that therein, above all, consists its great 
" utility." 

The Abb6 Sibire (who spoke for Nicholas Lupot) 
expressed a similar opinion (75). 

The concurrence of M. F^lix Savart with this view 
will, a little later, be quoted at length. 

The table of a violin (the back also in a somewhat 
auxiliary sense) is intended to be a diaphragm which, 
receiving its vibratory impulse firom the strings, 
through the bridge, communicates the vibrations to 
the air contained in the instrument. 

In order to fulfil the delicate functions required, 
such a diaphragm must possess a certain amount of 
homogeneity, its component fibres must not be sus- 
ceptible of independent motion or vibration ; it must 
be suflBciently elastic to respond to the slightest im- 
pulse, but its resistance must give rise to vigorous 
reaction. 

In trying to understand this complex problem, let 
an extreme case be supposed: a diaphragm is con- 
structed of unsized paper, i.e.y of practically pure 
cellulose in the form of interlaced, elongated cells. 
Vibration communicated to this diaphragm will be 
disseminated among the fibres of which it is com- 
posed, the motion will be individual instead of com- 
bined ; as a diaphragm it will be a failure. 

Saturate this diaphragm with a varnish which, 
when dry, will render it homogeneous and elastic, it 

V. D 



34 Varnishes of the Italian Violin-makers. 

will now respond to the most delicate vibration ; no 
impulse will be strong enough to cause individual 
motion among its fibres. The brilliancy of the tone 
produced will, of necessity, depend on the resistance 
which this elastic diaphragm offers to the vibrations 
which are communicated to it; for action and reaction 
are equal and opposite. This resistance will mainly 
depend on the physical attributes of the varnish with 
which the fibrous diaphragm is impregnated. 

Again, instead of. saturating the paper so 
thoroughly as to cause perfect homogeneity, let it 
be supposed that the varnish is diluted, in order 
that the impregnation may be partial, not complete, 
so far as the ultimate constituents of the varnish 
(minus the vehicle) are concerned ; the diaphragm 
will now vibrate, as when completely saturated, 
under the influence of a gentle impulse, but if the 
amplitude or repetition of the vibrations be carried 
beyond a certain point, the fibres now being only 
lightly agglutinated, an individual motion will be 
caused, as well as the combined one. This is, 
apparently, the effect intended to be described when 
an instrument is said to be overplayed. 

This latter supposed condition of the imaginary 
diaphragm represents roughly the finished, unvar- 
nished violin. It will be easily conceived that, if the 
fibres of the wood be only lightly agglutinated, any 
long-continued vibration will have a tendency to 
cause increasing dislocation (a slipping of one fibre 
on the other) — individual motion instead of combined 
movement. This dislocation may be permanent or 
temporary. 

If the old instruments be carefully examined, it 






The Tone of Violins. 35 

will be found that not only did the varnish cover the 
wood, but it saturated its outer layer, thus forming a 
homogeneous and elastic diaphragm such as has been 
described. The outer stratum of the wood (forming 
table, back, and ribs) is no longer soft and porous, it 
is homogeneous and compact ; in fact, after all the 
surface varnish has been worn away, the bare wood 
takes a fine polish ; it is clearly to be discerned that 
the varnish has penetrated the wood to some depth. 

Mr. Edward J. Payne, author of the article on 
Antonio Stradivari in Grove's " Dictionary of Music," 
writes thus of Stradivari's varnish: — " It is oil varnish 
^' of a soft and penetrating nature, apparently per- 
meating the wood to some depth beneath the sur- 
face, so that when the body of the varnish is worn 
" off the colour and substance appear to remain." 

Probably no one has made so many experiments 
on the tone of instruments as the late M. F61ix Savart, 
of Paris. His opinion is thus quoted by M. Mailand 
(46) (the translation is mine) : — " The varnish adds 
" to the beauty at the same time that it renders the 
" quality of sound permanent ; when the vamish- 
" ing of the table is neglected, the instrument loses 
" its mellowness and its force. No one is ignorant 
" that guitars whose tables are not varnished lose 
" much with age ; it is the same with pianofortes. It 
might be thought that in this latter instrument, the 
deterioration, after some years of service, would 
depend entirely on the considerable pressure 
" exerted on the table by the tension of the multi- 
" plicity of strings ; but it derives its origin from 
" another source. If it depended only on the cause 
" which has been assigned to it, the sound would not 

d2 



ti 



€1 



36 Varnishes of the Italian Violin-makers. 

" become acute or shrill (aigu) and meagre (maigre) 
" as it does become : it would become dull (sourd = 
" deaf) and feeble (faible), as I have had reason to 
** remark (or observe) that it does when I made violins 
" of which the tables were too thin, or to which I 
" fixed no bar ; they became dull (sourd) in propor- 
" tion to the sinking of the table, and never meagre. 
It appears that in pianofortes the vibrations com- 
municated to the table by a great number 'of strings 
" which vibrate harmonically with the principal 
" sound (prime note), destroy little by little the 
" texture (contexture) of the wood, with the expul- 
" sion of a great number of particles in the form of 
" dust ; for if the firwood which has been used in a 
" pianoforte is wrought, it appears to be very porous 
" and as if rotten. It is presumable that the humidity 
" of the air counts for a good deal in this change of 
its nature : for violins do not deteriorate, although 
much pressed by the tension of their strings, while 
" guitars, which are no more pressed, deteriorate 
" very promptly. It is the same with violins when 
" they are not varnished : the sound has at first more 
" mellowness and force than if they had been 
" varnished ; but it becomes modified little by little, 
" and at length feeble and meagre, which happens 
" in a tolerably short time." 

M. Savart is further quoted (58) : — " In general 
" the violins are most esteemed which are covered 
" with an oil varnish " {i,e,^ a varnish whose vehicle 
is oil of turpentine). " I believe well enough that this 
" is with reason, since it is more adhesive than that 
" which is made with spirit of wine ; it is more suit- 
" able for instruments of which the tables are thin, 






The Tone of Violins. 37 

" because by penetrating them, it gives them more 
" consistence. I think, on the contrary, that for 
" violins of which the tables are thick, a better 
" varnish is that which permeates less into the wood 
" and which leaves it all its natural qualities. That 
" of lac dissolved to saturation in rectified spirit of 
" wine of 34 or 36 degrees has appeared to me very 
" suitable ; it dries promptly and is not liable to scale 
" off." Speaking of colours, such as saffron and 
amatto, he says : — " It appears that they have a bad 
" effect on the wood, and contribute to make the 
" sound piercing or shrill (aigre) ; lac varnish gives 
*' a very beautiful tint, to which one might hold one- 
" self (i laquelle on pourrait se tenir), and which has 
" no bad effect whatever." 

M. Mailand thinks M. Savart's remarks as to 
the different treatment of thin and thick tables illo- 
gical. 

The question of choice of vehicle is of paramount 
importance. It is obviously useless to possess a suit- 
able resin if it simply remains on the surface of the 
wood, as it would do in the spirit varnish which M. 
Savart proposes ; if lac were soluble in oil of turpentine 
the position would be changed. The strong affinity 
which most woods possess for this latter vehicle is re- 
markable (not very easy to understand). The absorption 
of alcohol and water by wood is neither very rapid nor 
complete ; but if a section of wood be dropped into oil 
of turpentine, saturation appears to be instantaneous ; 
the wood at once becomes diaphanous, which is not the 
case when it has been immersed in the two other fluids 
for some time. Duration of evaporation is an equally 
important factor ; alcohol leaves resin so speedily that, 



38 Varnishes of the Italian Violin-makers. 

apart from affinity, no reasonable time is allowed for 
absorption. 

Let us, firstly, consider the question which M. 
Savart raises relating to climatic influences. 

It is somewhat remarkable that the relations of 
varnish and moisture to the wt)od of a violin have not 
been, ere now, exactly defined. Such questions may 
not be neglected, because eventually they may prove 
to be of great importance. When, during the course 
of these experiments, the time had arrived for testing 
varnish on a virgin violin ("in the white"), the first 
thing requisite was to accurately weigh the instru- 
ment. A coat of one of the oil-varnishes, hereafter 
described, was laid on with a brush ; as soon as the 
varnish was sufficiently dry, the violin was again 
weighed : it weighed rather less than it did originally. 
This result was certainly unexpected. The same ex- 
periment was repeated with another instrument — ^with 
the same result. One of two things must have hap- 
pened : — I St. The varnish had displaced from the 
wood something heavier than itself (? water) ; or 2nd. 
While the varnish was drying, the wood had lost 
more weight by elimination of moisture than it had 
gained by the coat of varnish. 

On close investigation it was found not to be so 
easy to determine the exact weight of an unvarnished 
violin as might be supposed ; the weight was never 
constant from one day to another. If the instrument 
was hung up in a room where the temperature is 
practically constant, at about 16° C. (60° F.), its 
weight changed with changes of weather which 
affected the hygroscopic condition of the air. 

The weight of a virgfin violin (without finger- 



The Tone of Violins. 39 

board), which has been made for some years, will 
generally be found to be between 4,400 and 4,900 
grains (say 285 to 317 grammes) at 60° F. This weight 
wilL increase or decrease to the extent of at least 
60 grains (4 grammes) when the violin is exposed to 
the air of different apartments in the same house. If 
the weight of the unvarnished instrument, when in an 
ordinary sitting-room, be taken as 4,500 grains (300 
grammes), it may be safely assumed that of this weight 
about 500 grains (33 grammes) are water. This con- 
siderable proportion of water present in wood (even 
when well dried and seasoned) makes it very difficult 
to determine with exactitude how much weight the 
virgin instrument gains by the varnishing of its 
exterior surfaces. 

It has been found that the increase of weight in a 
violin by the varnishing process ranges from about 
200 to 400 grains (13 to 26 grammes). The actual 
weight of the varnish will slightly exceed this, because 
it must be taken into account that the varnish which 
has permeated the wood has in all probability dis- 
placed an equal volume of water. 

If the wood of violins were varnished on all sides 
there might be more reason for the conclusion at 
which M. Savart arrived in endeavouring to explain 
the reason for the greater deterioration of the soimding- 
boards of guitars and pianofortes than of those of 
violihs. He appears to have overlooked the fact that 
the inside of a violin is unvarnished ; therefore, from 
this direction the wood will be always freely affected 
by the humidity of the atmosphere. 

It is easy to ascertain whether a varnished violin 
is so affected. A completely varnished, unmounted 



40 Varnishes of the Italian Violin-makers. 

violin {i.e.y one without finger-board, tail-piece, bridge, 
sound-post, pegs, &c.), which has been hanging for 
many months in an even temperature of about 60° F., 
and in an atmosphere which is comparatively dry, is 
accurately weighed ; its weight is 5,038 grains. It is 
removed to another apartment (in the same house) on 
an upper floor, where the temperature is about 10° F. 
less, the air more humid. 

After 24 hours its weight is 5,056 grains ; 

After 48 hours its weight is 5,068 grains. 
Returned to the warmer apartment it loses weight 
again until it reaches its normal of about 5,038 grains. 
Its weight will always depend upon the state of the 
external atmosphere. When frost sets in, its weight 
falls, in a few hours, below 5,000 grains ; on account 
of the dryness of the air of the room consequent on 
the raising of the temperature of the cold, dry, external 
air which is gradually introduced, to 60° F. 

I have made great numbers of careful and exact 
experiments on the question of the absorption of 
moisture by wood (new and old, seasoned and un- 
seasoned) from the atmosphere ; the weight of wood 
is never constant, it follows the atmospheric conditions. 
Even doors which are French-polished on both sides 
and on their edges modify their weight and dimen- 
sions with the changes of the weather and of the 
seasons. 

It must be evident, therefore, that M. Savart was 
labouring under soma misapprehension when he 
ascribed the differences in the deterioration of the 
tables of violins, guitars, and pianofortes to climatic 
influences. Manifestly, the explanation must be 
sought in some other direction. 



The Tone of Violins. 41 

Having disposed of the question of climatic in- 
fluences, we have, secondly, to consider the other 
point which M. Savart raises by his suggestion that 
in an unvarnished sounding-board the vibrations 
destroy, little by little, the texture of the wood. It 
must be admitted that this proposition is presented in 
a somewhat loose form, even if we substitute his own 
word " contexture " for texture. It is not quite clear 
what he means, or whether he had formed any definite 
mental conception. 

In order to obtain such an exact mental concep- 
tion of the possible eflfect of vibrations on the structure 
of wood, a somewhat long digression will be unavoid- 
able. 

The cells of which trees and plants are constructed 
may be described (for the present purpose) as vesicles 
having thin membraneous walls, formed of celliilose, 
but in nearly all cases more or less thickened by other 
substances. These membranes are permeable by 
fluids. The forms of these cells are of infinite variety ; 
they appear generally to be derived from spheres, 
spheroids, or tapered cylinders, the derived angular 
or irregular forms being probably dependent on the 
conditions to which the cells are subjected, according 
to their positions in various structures. 

Everyone knows that a vessel or vesicle formed of 
a thin membrane has vastly different properties 
whether it is filled with a very light and elastic fluid, 
such as air (uncompressed),' or with a comparatively 
dense and inelastic one, such as water; the substitution 
of the latter fluid for the former has a great stiffening 
effect on the vessel — it becomes much more rigid. In 
the case of the thin-walled cells of parenchyma, this 



42 Varnishes of the Italian Violin-makers. 

may be one of the reasons for increase of rigfidity, but 
it cannot apply to the sclerenchymatous cells of wood, 
which are much affected by accessions of moisture 
(such as absorption from the atmosphere) which are 
quite insufficient to fill the cells; in this case the 
increase in rigidity must be ascribed to that thickening 
of the cell-walls by the absorption of moisture which 
is known as turgidity. 

We have before us a succulent plant constructed 
of cells in various forms. The tapered stem is erect ; 
the branches are either horizontal or slightly removed 
from the horizontal position upwards or downwards 
or bending in graceful curves; attached to these 
branches, by appropriate stalks or petioles, are broad 
leaves expanded in planes approximately horizontal. 

Arrangements eire made to cut off the usual supply 
of water to the absorbent roots of the plant. Evapora- 
tion continues from the leaves until the water in the 
cells is dissipated in the surrounding air. The stem 
is now no longer able to support the weight of the 
organs attached to it — it bends and inclines to one 
side ; the branches decline from their usual position ; 
the leaves become flaccid and droop. 

Water is again supplied to the roots ; in a short 
time the plant resumes its original graceful form. 

These phenomena are, of course, due to the loss 
and recovery by the cells of that stiffness or rigidity 
which has been already ascribed to the presence of 
water. It must be evident, however, that if each 
individual cell occupied an isolated position no such 
effect could be produced ; it is essential that the cells 
should be firmly attached to each other by some 
agglutinating substance, sufficient to cause adhesion 



The Tone of Violins. 43 

of their parts in contact, in order that individual 
rigidity may afford the combined strength necessary 
to withstand the strains of flexure and torsion to 
which the various parts of the plant are subjected. 

The rope-maker talces a quantity of hemp, reduced, 
by suitable means, to long fibres composed of bundles 
of elongated cells attached to each other by some 
agglutinant. He twists numbers of these into the 
form of a cord. He again twists a number of these 
cords into a strand. From several of the strands, in 
a similar manner, he finally constructs the rope; 
which is, in fact, composed of a complicated system of 
fibrous spiral springs. If this new, dry rope be 
strained between two fixed points, the amount of its 
tension may be ascertained by weights or other 
equivalent arrangement. If the rope be now saturated 
with water, the tension which it exerts at the two fixed 
points will be largely increased. This phenomenon 
is again due to the stiffening or increased rigidity of 
the fibrous spiral springs caused by the water absorbed 
by the cells ; in short, to turgescence, each individual 
cell being suitably attached or agglutinated to its 
fellows by a substance naturally or artificially intro- 
duced for this purpose. 

Wood consists of numbers of elongated cells simi- 
larly attached to each other. If dry wood be saturated 
with water it will be found to change its form ; it will 
be considerably enlarged laterally but not appreciably 
longitudinally — in common language, it swells. As 
it dries it resumes its original dimensions. This 
phenomenon is once more to be attributed to turgidity 
— to the increase and decrease of the strength or 
rigidity of the tiny cells, firmly attached to each other 



44 Varnishes of the Italian Violin-makers. 

by the adhesion of parts of their membraneous 
walls. 

Now a still more remarkable phenomenon has to 
be observed: If the alternate saturation and desic- 
cation of the rope and of the wood be indefinitely- 
repeated, it will be foimd that the intensity of the 
effects described grows gradually less and less — ^the 
longitudinal contraction of the rope, the lateral ex- 
pansion of the wood, slowly but perceptibly, decreases; 
the rope becomes old, the wood becomes seasoned. 

How is this interesting phenomenon to be ex- 
plained ? Have the ultimate cells of the hemp and of 
the wood gradually lost the property of turgescence, 
of acquiring increased rigidity from the absorption of 
water ? This seems inconceivable : fortunately we 
possess means of proving, beyond all doubt, that such 
is not the case. We take parts of the old rope and of 
the seasoned wood; by suitable chemical means we 
deprive the cells, of which these substances are mainly 
composed, of all extraneous matter until membranes 
of practically pure cellulose remain. These purified 
cells are suspended in sufficient water to constitute a 
conveniently diffuse magma ; paper is made fi-om it 
by two different methods, the one process being known 
as hand-making, the other as machine-making. In 
the first process, the workman dips a rectangular form 
(covered on the under side with fine wire gauze) 
beneath the surface ; he raises it ; as the water escapes 
through the gauze, by a dexterous shake he causes 
the fibres to cross each other or interlace at the 
moment when they are forming a sheet of paper. In 
the second process, the mixture of cellulose and water 
is caused to flow, in a wide, thin stream, on to a 



The Tone of Violins. 46 

slowly, forward-moving sheet of wire gauze ; this hsis 
also a lateral shaking motion which, to a certain 
extent, interlaces the fibres at the moment when the 
water is leaving them; but this interlacing is less 
complete than in the first process, for the current of 
the stream has a tendency to cause all the fibres to 
float in a longitudinal direction — their longer axes 
tend to remain in the same plane as that in which the 
water and the gauze are moving. 

When the sheets of paper so made are sufficiently 
dry, they are impregnated with a resinous or colloid 
substance to an extent sufficient to cause mutual 
adherence of the fibres without producing an imper- 
vious integument. 

The hand-made paper will be (practically) equally 
strong in every direction ; the machine-made will be 
much stronger longitudinally than laterally. If both 
kinds of paper, when dry, be again moistened, they 
will expand ; the hand-made nearly equally in all 
directions, the machine-made much more laterally 
than longitudinally. If the alternate processes of 
moistening and drying be repeated, it will be found 
that the ratio of expansion slowly, but surely, 
decreases. 

It is therefore perfectly clear that the ultimate 
fibres of the hemp and wood have retained their 
original attributes unchanged from first to last. 

In the problem before us we had but two factors : 
— I St. Turgescence, the increase of rigidity caused by 
absorption of water ; 2nd. The aggregation of effect 
caused by the mutual adhesion of the cells. It is 
submitted that clear proof has been adduced that the 
first factor is a constant ; it follows, of necessity, that 



46 Varnishes of the Italian Violin-makers. 

the characteristic changes in the structures, formed 
by the aggregation of cells, must be caused by some 
alteration in the second factor — their mutual adhesion. 

The explanation of the observed phenomena can 
no longer be doubtful : — ^The movements of the indi- 
vidual fibres, caused by the strains set up by the 
alternate wetting and drying, must have, to some 
extent, modified or destroyed their mutual adhesion ; 
by constant repetition they have acquired accommoda- 
tion to the requirements of their environment, the 
fiiction of their surfaces in contact has become lessened 
by repeated strain or movement, they slip one on the 
other to the required extent. 

In the case of ropes which are repeatedly sub- 
jected to complete immersion and saturation, a part 
of the agglutinant may have been removed by the 
solvent action of water ; this cannot be the case with 
wood which is subjected only to continual changes 
in the hygroscopic condition of the atmosphere. 

The seasoning, in the case of ropes or of pieces of 
wood (such as oars) which are alternately completely 
saturated, then partially or completely dried, will be 
quite different (in degree) from that of wood under 
shelter, exposed only to atmospheric changes. The 
seasoning or immunity from change of form will 
apply only within the limits of moisture and desic- 
cation to which the fibrous structure has become 
accustomed. 

If, then, the slow movements, arising fi-om strains 
which are set up in wood by changes of hygroscopic 
conditions, are capable of gradually loosening the 
adhesion of the fibres to each other to a small extent, 
it may be expected that continual vibrations, more or 



The Tone of Violins. 47 

less violent, will produce similar results in an aggra- 
vated form. 

As wood which is of a (so-called) mild character, 
i,e.j which contains a minimum quantity of evenly 
distributed agglutinating matter, will season more 
quickly than wood which is saturated with some 
resinous or adhesive substance (often so unevenly 
distributed as to cause cracks and fissures before 
the gradual accommodation to strains has had time 
to be perfected), so may it be expected that the 
influence of vibrations, in causing dislocation, disin- 
tegration, or (to use the expression of M. Savart) 
destruction of contexture, will be in inverse propor- 
tion to the degree of agglutination. 

M. Savart states, as the result of actual experi- 
ments, that violins which are unvarnished soon 
deteriorate in tone. This fact appears to be confirmed 
by other observers ; to my knowledge, the correctness 
of this observation has never been disputed. 

When due consideration is given to all the ob- 
served facts, only one conclusion appears possible: 
that the deteriorating influence is vibration, which 
causes a kind of dislocation of the fibrous cells of the 
wood — ^they slip or move on each other without much 
firiction or adhesion — individual motion becomes 
more or less possible. When varnish penetrates the 
wood so as to produce a practically homogeneous 
diaphragm, this dislocating action of the vibrations 
is prevented. 

Most persons who are in the habit of playing on 
a stringed instrument with a bow will have observed 
that if a violin, viola, or violoncello has not been 
played on for some months or years the tone will 



48 Varnishes of the Italian Violin-makers. 

not be produced at first with the same facility as 
after some weeks of daily use ; the sounding-boards 
of the instrument will olBFer more resistance at the 
beginning than at the end of a period of frequent 
playing ; there will be a perceptible difference in the 
tone after a period of complete rest as compared with 
one of daily practice — always supposing that a good 
instrument is in question. The Messrs. Hill, writing 
on this subject, express themselves very strongly 
(Stradivari, 239): — "To close — one most earnest 
" word. Instruments by continual use are apt to 
" become weary. They may even virtually be killed. 
" Give them rests. We feel it a duty to urge most 
" strongly that fine instruments should not be brought 
" to premature death by ceaseless use." 

Physiologists have g^ven much study to the weari- 
ness, or decline of power of response, which overtakes 
living cells (animal and vegetable) under varying 
conditions of excitement and work. It can be no 
such problem which is set before us in considering 
the change in the condition of the dead cells of a 
stringed instrument caused by repeated and varied 
vibrations. This can be no question of exhausted 
vitality or of subtle electrical conditions — ^the change 
must be purely physical. 

During the course of these experiments an ^nci-^ 
dent occurred that throws some light on this subject. 
Certain violins were varnished with pure turpentine ' 
varnishes (containing no oil). After drying, for at 
least six weeks, it was found that if the surface of 
the varnish (which was not sticky) was briskly rubbed 
/with the fingers (a cloth would not do) the surface 
pulverised. If this powder was careftiUy removed, 



The Tom of Violins. 49 

by means of a dry cloth or brush, a dull, opaque 
surface of varnish would remain. After a few hours, 
or at most a day or two, this dull surface spontane- 
ously regained (temp. 60° F.) its former smooth, 
translucent, bright appearance. This experiment 
was repeated with various similar varnishes with 
the same result. 

The particles of varnish which had become dis- 
integrated by friction regained their cohesion after 
a period of rest. 

The belly or table of a violin is formed of a thin 
section of firwood (Abies or Picea). The elongated 
cells or fibres (interspersed with medullary rays) of 
which it is composed are agglutinated with a soft 
resin (terpene oxide), sparingly distributed. The 
proportion of this agglutinant is not enough to con- 
stitute with the sclerogen a diaphragm of sufficient 
homogeneity or endurance to resist permanently the 
disintegrating influence of strong musical vibrations ; 
when subjected to these, for any length of time, the 
quality of the sounding-board becomes deteriorated ; 
it does not regain its original condition after a period 
of rest : apparently the amount of resin present is 
insufficient to allow of re-cohesiori. 

If the exterior surface only of this sounding-board 
be coated with varnish of such a quality that it is 
able to permeate the wood to a greater or less depth, 
a very complex condition of this fibrous diaphragm 
must ensue ; the exterior surface of the wood will be 
(it maybe assumed) completely saturated, while the 
interior surface will not be reached by the varnish 
or in any way affected. Between these two extremes 
of complete saturation and total immunity there will 

V. E 



60 Varnishes of the Italian Violin-makers. 

be gradations of absorption, constituting zones which 
will offer gfreat or little resistance to the dislocating 
influence of vibration. If this assumption be well 
founded (the recorded facts seem to leave little room 
for doubt), it is not difficult to conceive what will 
happen. The wood which has received no accession 
of agglutinant from the varnish, after a comparatively 
brief period of use, will become permanently disquali- 
fied ; this disposes of the innermost zone. Next we 
have to consider the zone which has received from 
the varnish the minimum quantity of agglutinating 
material ; by constant vibration it may be supposed 
that this zone also, at a little later date, will become 
disintegrated, at least temporarily. A similar effect 
will ensue, after still longer endurance, in the zones 
with constantly increasing increments of varnish or 
resin (short of sufficient to constitute homogeneity), 
consequently with gradually augmented power of 
resistance to the dislocating influence. In this way, 
the effect of constant playing (continued vibration) 
must be to gradually diminish the thickness of the 
effective diaphragm. 

With rest (if a similar resumption of cohesion takes 
place within the wood as has been shown to ensue 
with the varnish on the surface), those zones, which 
have not entirely lost their power of recuperation {i.e,j 
where the amount of resin is sufficient to allow of 
re-cohesion), will soon resume their original condition 
of resistance : the violin will regain its normal state 
of sonorousness and brilliancy. 

It seems impossible to conceive any other hypo- 
thesis which will satisfy all the conditions. 

Practical experiments have confirmed these theo- 



The Tone of Violins. 51 



retical conclusions. There seems to be no doubt that 
not only does the production of satisfactory tone 
depend on the permeation of the wood by the elastic 
varnish with which the instrument is covered, but the 
particular quality of tone produced is influenced by 
the characteristics of the varnish which has been 
absorbed. 

A general impression appears to prevail that the 
tone of instruments of the violin family improves with 
age. The facts on which this opinion is based are 
not evident. The deterioration of the tone of some 
instruments by use and age seems to have been satis- 
factorily demonstrated, but evidence of the converse 
is wanting. It is admitted that the tone of many old 
instruments, especially those of Italian origin, is 
generally superior to that of modem instruments, but 
there appears to be no sufficient reason for supposing 
that with age the latter will become equal to or 
approach the quality of the former. It would appear 
that during the lifetime of Stradivari the superiority 
of his instruments was acknowledged and appreciated. 
There is no evidence that the artistes and amateurs 
who lived at that time, and who were quite capable of 
forming a correct judgment on the question, preferred 
the older instruments of Brescia and Cremona to those 
which calme direct from his hands (especially between 
1700 and 1725). 

Until real evidence is produced to the contrary, it 
would be more safe to assume that the old Italian 
instruments, which are so much esteemed to-day, owe 
their superiority to qualities which they have always 
possessed since they left the hands of their makers, 

e2 



62 Varnishes of the Italian Violin-makers. 

or which they acquired within a year or two from the 
date of their completion. 

Everyone who has studied Professor Helmholtz's 
admirable work on the " Sensations of Tone," will be 
aware that the distinctive qualities of tone, of human 
voices, and of various kinds of musical instruments, 
depend upon the strength of the upper partials which 
the ear analyses as accompanying the prime tones, 
thus forming compound tones made up of constituents 
of varying prominence. A thin string tuned to a 
certain pitch will produce a compound tone which will 
differ very considerably, in the prominence of its 
constituents, from that derived from a thick or covered 
string tuned to the same pitch. 

In violins the sympathy between the strings and 
the belly or table is so intimate that it is impossible 
to conceive that a change in the varnish, which is 
sufficient to influence the degree of elasticity or of 
resistance of the vibrating diaphragm — ^to cause a 
diminution or increase of amplitude of vibration — 
shall be without effect on the quality of tone. 



( 63 ) 



Chapter IV. 



ON THE MANUFACTURE OF OIL-VARNISH. 




jEFORE proceeding to attempt to describe the 
best known methods for the preparation of oil- 
varnish, it will be well to define as nearly as possible 
the attributes of a perfect varnish. 

A varnish consists of a soluble, transparent, 
adhesive solid, or of a combination of such sub- 
stances, dissolved in a volatile fluid which is called 
the " vehicle." When the vehicle evaporates, a solid 
pellicle remains on the surface to which the varnish 
has been applied. A very little consideration will 
make it evident that, if the best results are to be 
obtained, this pellicle must be perfectly homogeneous ; 
in other words, if it consists of more than one sub- 
stance, its constituents must be, not in a state of 
intimate mixture only, but blended — actually incor- 
porated the one with the other. 

Take an oil-varnish, for instance : Copal is fused 
and dissolved in linseed-oil (properly prepared for 
the purpose) ; this viscid solution is then diluted with 
the volatile vehicle (oil of turpentine), to render it 
convenient of application, by means of a brush. When 
the vehicle evaporates spontaneously, a thin and per- 
fectly homogeneous pellicle remains : the copal is 
dissolved in and incorporated with the linseed-oil. 
This is conceived to be a perfect varnish. 



54 Varnishes of the Italian Violin-makers. 

Suppose it is desired to give to an oil-vamish 
an artificial colour, by means of some substance which 
is insoluble (or only partially soluble) in linseed-oil 
or oil of turpentine. The pigment is dissolved in 
alcohol; this coloured solution is intimately mixed 
with the varnish. When the alcohol and the oil of 
turpentine evaporate, the colouring-matter is pre- 
cipitated in a state of fine division in the varnish, but 
the pellicle is no longer homogeneous ; although the 
colouring-matter may be distributed ever so evenly 
or skilfiilly, it is adventitious, foreign (so to speak) to 
the other and main constituents of the pellicle ; it is 
not dissolved in nor incorporated with either of them. 
It is suggested that a varnish so constituted is mani- 
festly imperfect. 

No one who studies the existing descriptions of 
the manufacture of oil-varnishes can fail to be im- 
pressed by the surprising technical skill, requisite for 
their successfiil production, which has been for so 
long a time available. The maker of oil-vamish is 
an expert : a man of high intelligence, quick ob- 
servation, cool judgment ; prepared with ready wit 
and presence of mind in all emergencies. But a very 
little consideration will show that this important 
industry rests more upon the skill and experience 
which have grown up firom long practice than upon a 
scientific basis or exact knowledge of cause and effect. 

Among the many writers on this subject, firom 
Professor Tingry to those of the present time, no 
one appears to have described the processes with 
more lucidity and exactness than M. Laurent Naudin. 
His excellent treatise forms one of the numbers of 
that admirable series of technical works known as the 



On the Manufacture of Oil-varnish. 55 



ii. 



Encyclop6die scientifique des aide-m6moire/' pub- 
lished under the direction of M. L6aut6, Membre de 
rinstitut. 

His description of the preparation of a copal oil- 
varnish may be thus translated and somewhat 
abridged: — "The pieces of copal are selected for 
** colour and hardness. The great point is to select, 
" as nearly as possible, pieces of the gum which have 
" the same degree of fusibility. These selected frag- 
** ments are cleaned and dried ; they are then ready 
" for fusion." 

" The softer (demi-dur) copal produces varnishes 
** less coloured, but less resistant, than the hard copal 
« (copal-dur)." (31.) 

The temperatures at which these gums fuse are not 
very easy to arrive at ; the statements made on this 
point are somewhat confusing, if not conflicting. 
This is, however, not surprising, for the determination 
of the fiising-point of substances of this kind is per- 
plexing to most persons of moderate ability. These 
resins are very bad conductors of heat. Take an 
ounce of colophony (oxidized in a way which will be 
hereafter described), for instance, put it in a deep 
porcelain basin, heat the basin over a sand-bath. 
Soon the part of the resin in contact with the basin 
will fuse, a semi-fluid film will adhere to the surface 
of the basin. Now the lump of resin may be left at 
rest or turned and kneaded, but it will be found im- 
possible to cause the parts of the resin not in actual 
contact with the basin to approximate, in tempera- 
ture, to the parts which are directly heated. 

When, therefore, 360° C. is given as the tempera- 
ture necessary for the fusion of hard copal, 230° as 



it 
a 



66 Varnishes of the Italian Vtolin-makers. 

that for soft copal (demi-dur), it is understood that 
these are the temperatures of the interior surface of 
the vessel containing the resin or of the gas or vapour 
within it. The temperature of a mass of several 
pounds of copal can never, even when fused, approach 
uniformity. 

M. Naudin continues his description : — " The 
" operator places in a vase-shaped copper vessel, 
" with a flat bottom, about 4 kilogrammes of hard 
" copal. The copper vessel is then put over a furnace 
" with a brisk fire. The mass crepitates, water is 
" given off" (N.B. the resin was previously carefully 
dried), " the resin fuses. To prevent adhesion one 
stirs continually with an iron spatula. The water 
having been driven off, pungent vapours of a ruddy 
" colour rise from the orifice of the vase. When all 
" the resin is fused, which one ascertains by testing 
" the mass with the spatula, 1,500 to 2,000 grammes 
" of boiled linseed-oil, heated to 150° C, are poured 
" into the vase and its contents are well stirred to 
" favour solution." " Supposing all to have gone 
well this compound of linseed-oil and copal is 
diluted with oil of turpentine at once, while it is 
" still hot, to bring it to the necessary state of fluidity 
" for a varnish." 

But if too little oil of turpentine be added, if the 
varnish is too thick — and this defect is discovered after 
the varnish has cooled — it cannot then be further 
diluted ; if cold oil of turpentine be added to the cool 
varnish, precipitation takes place, and the varnish is 
spoiled. In fact, it would appear that, after the 
varnish has once been completed and allowed to cool, 
there is no known method by which its fluidity may 






On the Manufacture of Oil-varnish. 57 

be increased by adding oil of turpentine, hot or 
cold. 

This is the case only with hard copal made into a 
varnish in the way described ; it would appear that a 
varnish made with soft copal, at a comparatively low 
temperature, may be diluted at pleasure. 

The production of an oil-varnish from hard copal 
can be effected only in the way described. At first 
sight, it would appear to be a very strange proceeding 
to expose a quantity of resin or gum (which, as already 
explained, is a very bad conductor of heat) to the 
action of a brisk fire, in a copper vessel, until the * 
temperature reaches 360° C, and then, when the resin 
fuses at this high temperature, to add linseed-oil at 
150° C, in order to effect its solution. The fluid oil, 
by conduction and convection, conveys heat much 
better than the solid, bad-conducting resin can possibly 
do. Why not pulverize the resin, mix it intimately, 
in the desired proportions, with the oil and subject 
this intimate mixture to the necessary degree of heat 
to induce solution ? Or, more simple still, why not 
mix together the powdered resin, the linseed-oil and 
the oil of turpentine, in their relatively correct propor- 
tions to form a varnish, raising this mixture to 360° C. 
(if this temperature is really necessary) to promote 
solution ? All these plans have been tried and have 
been found to be unsuccessfiil. 

M. Naudin describes the experiments which have 
been made in this direction : — ist. The proposal of 
M. P. Schiitzenberger in 1856 to effect combination 
of gum, oil, and essence in an autoclave at 300° C, 
duration 2 hours ; 2nd. A new attempt of M. H. Violette 



68 Varnishes of the Italian Vtolin-^nakers. 



in 1866 with the same object. Temperature 350-400°; 
pressure about 20 atmospheres. 

The information which M. Naudin gives as to the 
chemistry of copal and the eflFects which heat produces 
on this substance is sufficient to explain the causes of 
failure, at least in a great measure. He writes (28) : — 
" The most oxygenated resins are the most soluble." 
" Filhol has demonstrated, besides, that if the hard, 
" half-hard, and soft copals shewed notable diflFerences 
" in their physical properties, their centesimal com- 
" position was very nearly the same." 

As to the effects of heat (29) : — "It is only by 
" depolymerizing it towards 360°, by naked fire, that 
" one succeeds in overcoming its insolubility. Liquid 
" volatile products (isomers of essence of turpentine 
" C^® H^^) are thus obtained together with modified 
" copal (pyrocopal). But it must be distinctly said 
" that this new property is obtained only by changing 
" the properties of the gum. Tingry acknowledged 
" this by saying, * The fire profoundly changes the 
" * copal, and leaves it neither the elasticity nor the 
" * tenacity which it possesses in its natural state.' " 

"As early as i860, Schiller had extracted an 
" isomer of turpentine fi-om copal. A similar fact had 
" been observed by Dcepping with amber." 

" More recently Schwarz has made an elementary 
" analysis of pyrocopal ; fi-om which he concludes 
" that the body resulting firom the action of naked 
" fire is richer in carbon and poorer in hydrogen and 
" oxygen than the initial substance. In fact, one 
" always finds a very considerable quantity of water 
" in the crude oil of copal." 

From this information it is not difficult to deduce 



_ I 

On the Manufacture of Oil-varnish. 59 

some of the reasons why the eitperiments to which 
reference has been made have failed, and must always 
fail : — I St. In order to be soluble the copal molecule 
must retain a certain amount of oxygen. 2nd. The 
first effect of heat is the splitting of the molecule into 
a hydrocarbon and an oxide. 3rd. If the heat con- 
tinues, the oxide becomes dehydrated ; if dehydration 
is carried too far insolubility ensues. 

The apparent object of the operator in the ordinary 
process is to fuse the resin as rapidly as possible and 
to effect solution in the oil before extensive dehydra- 
tion has time to take place. The temperature of the 
oil being only 150° C, its admixture rapidly reduces 

I 

that of the copper vessel and its contents ; as soon as 
possible after solution, dilution with oil of turpentine 
still ftulher reduces the temperature. These apparently 
necessary conditions are not complied with in the 
autoclave. 

The effect of 360*^ C. on the copal is evidently not 
fusion, pure and simple ; the ultimate product (the 
volatile hydrocarbon excepted) is obviously a dehy- 
drated oxide (a colophone). 

But, apart from the causes of failure which have 
been suggested, it must be observed that linseed-oil 
has an afl&nity for oxygen which is somewhat greater 
than that of the terpenes (at any rate, when little 
degraded from the hydrocarbons) ; further, it will 
hereafter be demonstrated that when oil of turpentine 
is heated (even at 100° C.) with the other constituents 
of a varnish it also has a tendency towards important 
changes of constitution. Herein lies the probable 
explanation of the difficulties attending the ftirther 



60 Varnishes of the Italian Violin-makers. 

dilution of a varnish after it has been completed and 
cooled. 

The eflfect, therefore, of heating together copal, 
linseed-oil and oil of turpentine to so great an extent 
as 300° C. and upwards, must be exceedingly complex, 
owing not only to the changes in the copal, but also 
to the reaction of the three substances on each other ; 
with special reference to the oxygen which they 
contain and their varying affinity for it under the 
conditions to which the mixture is exposed. 

In the experiments which have been made it has 
been kept steadily in view that a clear understanding 
of these phenomena is of paramount importance in 
the production of varnish of the highest quality. 



( 61 ) 



Chapter V. 

OIL-VARNISH FROM TURPENTINE DERIVATIVES AND 

ITS APPLICATION. 




^HE simple observation of the dichroism of the old 
varnishes would have afforded little indication 
of their actual composition had not this been soon 
followed by an additional clue. 

On the 14th June, 1900, the brilliant morning light 
which streamed through an eastern window fell upon 
a panel of pitch-pine, not only revealing dichroism, 
but vividly recalling the two distinctive colours of the 
red Neapolitan varnish which had been the object of 
scrutiny more than a month before. 

It is deplorable how much one sees and how little 
one observes ! Every day, for many years, panels of 
pitch-pine had been round about me ; their admirable 
colour and figure (constantly changing according to 
direction, intensity and quality of light) could not be 
ignored; but, until now, I had failed to observe in 
their ever-varying shades a striking resemblance to 
the infinite colour variety of the old varnishes. Some 
of these panels had been covered with varnishes of 
different qualities : exposed to light and air for many 
years, the constituents of the wood had become slowly 
oxidized ; others had been oxidized by artificial 
means, devised to imitate these effects of time ; in 



62 Varnishes of the Italian Violin-^makers. 

many cases, the results of these two processes, natural 
and artificial, are incapable of distinction. 

The artificial process was very simple; it was 
varied in three ways: — ist. The finished wood was 
lightly coated with ordinary nitric acid; in a short 
time spontaneous oxidation ensued. 2nd. The nitric 
acid was diluted with an equal volume of water ; after 
its application, heat was necessary to induce a re- 
action. 3rd. The surface of the wood was saturated 
(more or less) with linseed-oil previous to oxidation 
(the panel which resembled the Neapolitan varnish 
was known to have been so treated). In all cases, 
after the completion of the oxidation process, the 
panel was washed with water and, when perfectly 
dry, was polished with alcoholic solution of shellac 
(French polish). 

The colours produced by these three variations of 
the oxidation process may be thus briefly described 
in general terms : those of the first and of the natural 
effects of time, transmission — ^yellow to yellow-brown ; 
refi-action — light red, deep orange, red-brown. Those 
of the second method are similar to those of the first, 
but tend more and more towards brown according to 
the degree and duration of the heat employed to 
induce the oxidation reaction. The colours of the 
third variety are deeper and more crude : the ground 
colour is yellow, buff, or orange ; that of refi-action is 
red (sometimes approaching crimson) and red-brown. 

It will be evident that in attempting to describe 
these colours of the wood all the different charac- 
teristics of the old varnishes of Brescia, Cremona, 
Venice, Naples, &c., have been portrayed. 

The term " pitch-pine " seems to have been 



OU'Varnish from Turpentine Derivatives. 63 

originally used to designate the wood of pinus 
rigida; but the fine wood of commerce which bears 
this name to-day is doubtless mainly the produce 
of pinus palustris, with the occasional unwelcome 
intrusion of that of pinus taeda, a timber of vastly 
inferior quality. 

The prevailing and characteristic property of this 
wood (as its name denotes) is its almost complete 
saturation with turpentine products. Pitch-pine is, 
in general, very elastic. 

A few weeks' study and investigation demon- 
strated the fact (naturally surmised) that the dichroism 
and colour effects which have been described are 
due to the presence of terpenes in different stages 
of oxidation, more or less dehydrated. 

Wherever the cradle of the violin may have been, 
it may be safely assumed that a bit of resin was in it, 
as surely as this substance is to be found in every 
fiddle-case to-day. The geographical position of 
Brescia indicates it as a most likely place in which 
to seek resin, as well as other products of conifers. 
It lies on the western shore of the Lago di Garda, 
whose opposite shore, especially towards the north, 
approaches the pine-clad hills of Austrian T3rrol. 

The surroundings of Venice are widely different 
fi-om those of Brescia ; but, if we may judge firom the 
fact that turpentine has been an article of commerce 
fi-om that city from time immemorial, larch-trees must 
abound at no very great distance. 

It is of interest to notice that although larch-wood 
is elastic, it is appreciably less so than is pitch-pine. 

It has already been demonstrated by the researches 
of M. Mailand, the results of which have been quoted, 



64 Varnishes of the Italian Violin-makers. 

that the substances which were most used in Italy in 
the sixteenth century for making varnish were de- 
rived from the coniferae of their own country. In 
fact, Alexis (1550) suggests only tw^o other substances 
as constituents of oil-varnish, viz., mastic and san- 
darac (what petrticular gum he designated by this 
name is problematical). 

Instead, therefore, of the old assumption that the 
violin-makers procured some remarkable gum (at 
present imknown, always rather mystical) for the 
basis of their varnish, it will be assumed that they 
used substances which they certainly had near at 
hand, both cheap and abundant. It will further 
require no great stretch of the imagination to suppose 
that the Venetians had mainly at their disposal the 
products of larch (Larix Europaea), while the Bres- 
cians and Cremonese had greater abundance of those 
of the pines and fir-trees (Pinus, Abies, and Picea). 

The turpentines most easily available to-day are 
resin and oil of turpentine from America, the produce 
of several varieties of pine, and Venice turpentine 
from larch. Ordinary commercial samples have been 
used in the experiments to be described ; their origin 
is unfortunately but unavoidably unknown ; their 
quality has been found to be good, invariably suited 
to the purpose for which they were procured. 

Henceforth a distinction will be made which is 
purely arbitrary: varnishes derived from resin and 
oil of turpentine (which are looked upon as products 
of pinus) will be called Cremonese; while, on the 
other hand, those derived from Venice turpentine (a 
larch product) will be called Venetian. Neapolitan 
varnish might have as its terpene constituent .either 



Oil-varnish from Turpentine Detivatives. 66 

the one form or the other ; the quality which charac- 
terizes it depends upon its mode of preparation, which 
will be hereafter explained. 

The drying oil used will be assumed to be linseed- 
oil ; this is the oil which the old authors on varnish 
appear to have known. Cremona, especially, appears 
to have gained a reputation for the excellent quality 
of its linen ; flax must have grown in the adjacent 
country, as in other parts of Italy, under conditions 
perfectly suited to its successful cultivation. 

The choice of an oxidant for the terpenes was a 
matter that required a good deal of consideration and 
some experiments. Not the slightest trace of any 
oxidant is to be found among the authors cited by 
M. Mailand, or elsewhere. It is conceivable that, at 
any rate in the earliest times, the old varnish-makers 
may have availed themselves of the well-known oxidi- 
zing powers of the essence ; but if so, even with the 
advantage of the warmth and sunlight of Italy, it must 
have been a slow process indeed. Life is too short to 
think of it as an available oxidant in this temperate 
and comparatively sunless climate ! 

Nitric acid was finally selected, and has proved to 
be well adapted in every respect for the purpose. 
Whether the reduction of the nitrogen is complete or 
only partial, the gaseous and aqueous residues are 
readily and entirely eliminated by heat. Its precise 
mode of application was less easy to determine. The 
method employed for the production of terephthalic 
acid was rejected as too energetic and selective, 
besides being wasteftil ; that eventually adopted will 
sufficiently appear from the experiments now to be 
described. 

V. F 



66 Varnishes of the Italian Violin-makers. 



If to icx) parts of oil of turpentine there be added 
about 50 parts of colophony or of Venice turpentine^ 
the whole warmed and mixed, a fluid results of con- 
venient consistency for heating in a glass, tubulated 
retort. With this viscous fluid, 3 to 5 parts of ordinary 
pure nitric acid (sp. gr. about i'42o), added drop by 
drop, are intimately mixed by stirring. On gently 
heating this mixture (over a water-bath) a reaction 
will begin, the violence of which may be moderated 
by removing the source of heat for a time. When 
efiervescence abates, the retort may be again heated, 
for some hours, until the reaction is practically at an 
end. When cold, more nitric acid is added, the same 
process being repeated. This process of nitrification 
and oxidation is several times repeated ; the resulting 
varnish is finally heated over the water-bath (at 
100° C.) for many hours. 

During the progress of this experiment the 
following phenomena may be observed. At first the 
nitric acid has a tendency to fall to the bottom of 
the retort, but by continued mixing the turpentine 
becomes turbid and of a brown colour. When the 
oxidation reaction begins, the turpentine soon passes 
to a yellow colour, which deepens as the oxidation 
proceeds ; all turbidity disappears. Seen in bulk the 
fluid passes firom orange to red; soon /only red rays 
pass through it ; at first, the red rays of a reading- 
lamp can easily permeate three inches ; but as the 
experiment proceeds it becomes less and less trans- 
lucent, although it is not turbid. After heating for 
several hours, when the nitric acid has been, appar- 
ently, long since all decomposed, bubbles of vapour 
or gas continue to rise through the fluid. A little 



J:.k«r<^e/l.*'^;Vs Uj;5-f^^ c^khf^*^-^ 

Oil-varnish from Turpentine Derivatives. 67 

volatile oil distils over into the receiver, accompanied 
always by water having an acid reaction and an odour 
of acetic acid. 

This 'volatile oil has a specific gravity of o*86i 
at i8° C. Its boiling point is not constant, but is 
about 155° to 156°. It is optically inactive. (The 
oil of turpentine used had sp. gr. 0*869 ^^ ^7i° C. and 
was strongly dextro-rotatory.) This oil therefore 
resembles terebene, which has the same odour. 

When cool, the varnish in the retort is strongly 
dichroic. It is decanted. At the bottom and near 
the sides of the retort a small quantity of red resin 
is found, which is insoluble in oil of turpentine, fireely 
soluble in aqueous alcohol, also soluble in warm 
linseed-oil. 

At the bottom of the retort there is a minute 
quantity of watery fluid ; when this is mixed with a 
little absolute alcohol, crystals of an organic acid 
immediately separate ; they possess the characteristics 
of succinic acid (as far as can be judged firom so small 
a quantity). 

The fluid decanted is a turpentine-varnish pos- 
sessing the desired qualities of colour and dichroism. 
The red resin found in the retort clearly indicates 
that such a varnish has its limits as a solvent. The 
insolubility of this part of the oxidized terpene was 
found to arise firom dehydration, caused by contact 
with the heated glass of the retort. From the fact 
that this red substance is readily soluble in linseed- 
oil, it may be inferred that the red varnishes were 
certainly oil-varnish. 

Turpentine-varnish produced in this way, fi-om 
either colophony or Venice tiirpentine, is of the 

f2 



I 



68 Varnishes of the Italian Violin-makers. 

highest quality, as far as transparency, colour, and 
brilliancy are concerned. It dries very slowly, but in 
a month or six weeks a pellicle of the usual thickness 
seems to become perfectly dry. If it is then rubbed 
briskly with the fingers, the surface pulverizes. If 
this powder is removed by a cloth or brush, the 
I surface which remains (only a small pgurt of the 
varnish has been removed) is dull and opaque ; but 
in a few hours it regains its former brilliance and 
translucence. The cohesion destroyed by the friction 
of the skin is restored after a period of rest. 

A thin pellicle of this same varnish applied to a 
door has borne ordinary wear for some years. It has 
been also found that thicker pellicles have slowly 
become quite solid. If such a varnish were made 
over a charcoal stove or similar directly-heating 
apparatus, it would precisely resemble the old var- 
nishes of Brescia. The pellicle would be entirely 
soluble in alcohol, and therefore when dry could not 
be distinguished firom that formed firom a spirit 
varnish. This is a probable explanation of the state- 
ment by M. F6tis that Magg^ni*s were spirit varnishes ; 
a spirit varnish at the time of Magg^ni seems to be an 
anachronism. 

After some few practical experiments with these 
simple turpentine-varnishes, it became evident that 
the be^t remedy for this want of cohesion was the 
addition of a certain proportion of linseed-oil (boiled). 
The old makers appear to have been led to the same 
conclusion. There are no indications as to the 
amount which they used; but, from the evidence of 
the varnishes, it is apparent that it was restricted to 
an approjpmate minimum. The formulae of Watin 



OiUvamish from Turpentine Derivatives. 69 

and Martin, given by M. Naudin, have been useful 
guides ; the results have proved satisfactory. 

An oil-varnish cannot, however, be properly pre- 
pared by simply adding linseed-oil to a turpentine-) 
varnish; it requires a special mode of manufacture, 
which will be now described. Firstly, what may be 
called the " dry way," will be given : — 

Colophony (such as is used for common purposes, 
and is found in commerce as rosin) is pulverized, in a 
mortar of Wedgwood porcelain, with a pestle of the 
same material. To this fine powder is added 20 to 
25 per cent, of nitric acid. The acid is added drop by 
drop (firom a pipette), and is quickly kneaded and 
mixed with the colophony, by the pestle, so as to 
prevent, as far as possible, unequal action. If this 
operation is deftly performed, very little rise of tem- 
perature will take place ; the result will be a rather 
dark, olive-green powder, somewhat granulated, which 
is quite stable at atmospheric temperatures. This 
nitro-colophone is, to this extent, a mild explosive : it 
requires a temperature approaching 100° C. to cause 
its decomposition ; when this begins, heat is generated 
by the reaction, consequently, decomposition proceeds 
with constantly increasing speed as the temperature 
rises, until, if the quantity dealt with be large, a 
violent reaction will eventually ensue. 

It is difficult for the gas and vapour evolved to 
escape from the viscous mass, which, as soon as the 
reaction commences, swells and froths; it is neces- 
sary, on this account, to conduct the operation in a 
vessel (preferably a deep porcelain basin) of sufficient 
capacity to retain the frothing resin. 

To excite the reaction the vessel containing the 



70 Varnishes of the Italian Violin^^makers. 

nitro-colophone may be heated over a water-bath, the 
temperature of which is quite sufficient for the purpose. 

If an excess of nitric acid be allowed to remain in 
contact with a part of the colophony, a reaction will 
there commence and spread to the whole (without 
extraneous heating). 

During the reaction the hot resin may be stirred, 
with a glass or porcelain rod or spatula, to facilitate 
the escape of steam, nitrogen and some of its lower 
oxides ; the heat of the bath should be maintained 
for a sufficient time (half-an-hour to an hour) until 
the resin has been practically freed from these pro- 
ducts of decomposition. 

The colophony, thus oxidized by the reduction of 
the nitrogen, will be found to have undergone a com- 
plete change, chemical and physical. Its colour, if 
the proportion of nitric acid (effectively used) has been 
at least 25 per cent, of the colophony, will be a bright, 
pure yellow (very much resembling gamboge) ; if the 
proportion of nitric acid has been considerably less 
than 25 per cent., the colour of the product will be 
less pure, inclining to brown. Its fusing point will 
be raised above that of the original colophony. 

Parenthetically, it may be here remarked that the 
fusing point (135° C.) usually attributed to colophony 
appears to be considerably too high (the difficulty 
attending such determinations with these very bad 
conductors of heat has been already alluded to). If a 
little finely-pulverized colophony be shaken on to the 
surface of boiling water, the powder will rapidly melt 
into oily globules. If a quantity of this powder be 
placed, in the form of a cone, in a porcelain basin and 
be exposed to the temperature of a water-oven (about 



Oil-varnish from Turpentine Derivatives, 71 

97"^ C), for a short time, it will fuse into a clear, 
homogeneous, viscid substance having a horizontal 
surface and every other attribute of a fluid. In 
linseed-oil and in oil of turpentine colophony fuses 
at the temperature of the water-bath ; consequently, 
the heat of this apparatus has been always found to 
be suflGLcient to aid its speedy solution in these 
menstrua. It seems, therefore, that the fusing-point 
of colophony cannot be above ioo° C. 

The mode of oxidation of colophony just described 
is not very convenient (except, perhaps, where the 
object in view is the production of spirit-vamish), 
because it is not always easy to ensure an even 
mixture of the nitric acid before oxidation begins ; it 
may easily happen that a few drops of nitric acid 
may, inadvertently, have time to attack a small 
portion of the colophony, raising its temperature suffi- 
ciently to initiate a reaction, before general and 
imiform nitrification has taken place. 

The subsequent steps in the formation of an oil- 
varnish from this bxy-colophone will sufficiently 
appear in the treatment of similar substances pro- 
duced by the " wet way " of oxidizing, which is more 
convenient and effective, and which will now be 
described. 

Colophony is dissolved in about half its weight of 
oil of turpentine, which is sufficient to form with it a 
viscid fluid, convenient for stirring. With this the 
necessary proportion of nitric acid is blended, the 
acid being cautiously added, in small quantities at a 
time (preferably from a dropping pipette or bottle), 
imtil, by continued stirring, an even mixture or com- 
pound is obtained. If during this operation the tern- 



I 

72 Varnishes of the Italian Violin-makers. 

perature should rise more than a few degrees, the 
containing vessel should be cooled by immersion in 
water. 

Venice turpentine generally furnishes, in its usual 
viscous state, a suitable substance for the process just 
described; if a little too thick, it may be easily 
thinned by admixture of a little essence. 

The nitrification of oil of turpentine is easily 
effected by this method. 

The object of the operator should be to ensure 
complete nitrification before proceeding to decom- 
position. The progress of nitrification may be easily 
observed by the change of colour of the terpene, 
combination being always attended by deepening 
brown. 

A little experience will show that the farther the 
terpene has been removed (by degradation) from the 
hydrocarbon the more rapidly it will be attacked by 
the nitric acid ; nitro-colophone will be formed in a 
few hours, while nitro-pinene (from oil of turpentine) 
will need as many days. As a clear understanding of 
this matter is of importance, it will be necessary here 
to describe in detail the preparation of the latter 
substance. 

A quantity of well-rectified, perfectly colourless, 
dextro-rotatory oil of turpentine (^-pinene) is placed 
in a deep porcelain basin (of a capacity largely ex- 
ceeding the volume of the fluid) ; ten per cent, (by 
volume) of pure nitric acid (sp. gr. 1*420) is added to 
it, drop by drop, well mixed by stirring with a spatula 
(glass or porcelain). Slight rise of temperature takes 
place, which may, if necessary, be controlled as 
already directed. The turbid mixture, on standing. 



Oil-varnish from Turpentine Derivatives. 73 

divides into two strata : the upper (terpene) colour- 
less; the lower (residue of nitric acid) red-brown. 
After standing for a few hours, a further ten per cent, 
of nitric acid is added, as before; on dividing, the 
upper stratum remains, apparently, unchanged. After 
standing for two days, the upper stratum has become 
of a pale brown ; the red-brown colour of the lower is 
fading. A fiirther ten per cent, of nitric acid is added 
precisely as before. Rise of temperature is now 
scarcely perceptible ; on dividing, the upper layer has 
become of a deeper brown. The characteristic odour 
of oil of turpentine has disappeared and an odour of 
fir-wood is substituted. The mixture is allowed to 
stand for two more days ; at the end of this time, the 
upper stratum has become of a dark-brown colour, 
the lower is almost colourless. Nitrification of the 
terpene is now supposed to be practically complete. 

It will be obvious, firom what has been described, 
that in the nitrification of a complex solution, such as 
one formed of colophony and oil of turpentine, or 
Venice turpentine alone (which is evidently a solution 
of isomerides and oxides in essential oil), the effect 
will be quite different whether a short or a long time 
is allowed for the process : the nitric anhydride will 
combine first of all with dehydrated oxides, later with 
oxides, and finally with the hydrocarbon itself. 
Nitrification may be confined to the action of a few 
hours, or may be extended to ten or twenty days, 
according to the intentions of the operator as to the 
constitution of the resinous part of his varnish. 

It will be remembered that the pitch-pine panel, 
whose colours were observed to correspond with those 
of an old Neapolitan varnish, was known to have been 



74 Varnishes of the Italian Violin-makers. 

saturated with linseed-oil previous to oxidation. It 
is easily surmised from this that the Neapolitan types 
were produced in this way: Some vamish-maker 
(here again I suspect the influence of some chemical 
student of the Church of Rome) obsCTved that the 
transformation of raw drying-oil into the more sicca- 
tive form is precisely on all-fours with the oxidation 
of turpentine. The old process was modified in 
accordance with this view. 

The first stage in the preparation of a Neapolitan 
varnish is this: Colophony or Venice turpentine is 
dissolved in raw linseed-oil (by means of a water-bath) 
in proportions suitable for the constitution of an oil- 
varnish, say about two parts of terpene to one part of 
oil. This oleo-terpene is nitrified in precisely the 
same way as that already prescribed for the terpenes 
alone. The disadvantage of the Neapolitan method 
is that the affinities of the two substances (turpentine 
and oil), acted upon together, are not quite the same ; 
consequently in the different stages of the preparation 
of the varnish, the transformations do not proceed 
pari passu (they tend rather in the opposite direction to 
that which is desired, as will be presently explained) ; 
in this case, nitrification should not be prolonged. 

Having thus prepared nitro-terpene or nitro-oleo- 
terpene, the succeeding stage — oxidation, by the 
decomposition of the nitro-compound — ^may proceed. 

The most convenient and safe method of performing 
this operation is to place over a boiling water-bath a 
capacious empty porcelain basin ; to this transfer the 
nitro-compound in suitable successive portions until 
the whole has been decomposed. If it is desired to 
attempt this transformation in the same basin in 



Oil-varnish from Turpentine Derivatives. 75 

which nitrification has been effected (which is prac- 
ticable after a little experience), the best plan is to 
warm the basin on one side only, so as to initiate the 
reaction at a point near the surface of the mass ; in 
this way, by a little management and care, decomposi- 
tion may be gradually extended to the whole mass. 

The oxidized resin or oleo-resin is allowed to 
remain over the water-bath until decomposition is 
quite complete and until the residual water of the 
nitric acid and its gaseous products have been driven 
off, this object being assisted by constant stirring; 
this may be generally attained in about an hour. 

The oleo-resin (for Neapolitan varnish), while still 
hot) is diluted with oil of turpentine, as desired. 

A requisite quantity of boiled linseed-oil is added 
to the oxy-terpene, in which it soon dissolves, the 
temperature of the bath being well maintained ; the 
resulting thick varnish may be then diluted, by the 
gradual admixture of oil of turpentine, to any extent 
that may be desired. 

All the varnishes thus produced are of a yellow 
colour, of varying depth : the colours of the Cremonese, 
Venetian, and Neapolitan varieties are somewhat 
different ; by persons well acquainted with their dis- 
tinguishing tints they may generally be identified 
without much difficulty. They are very transparent, 
perfectly bright and clear ; owing to their powers of 
refi-action and marked dichroism, they exhibit per- 
fectly the grain and every fibre of the wood on which 
they are laid. 

In the case of these varnishes, the addition of 
linseed-oil is not a necessity for the purpose of 
solution; for the resin, provided oxidation and sub- 



-'■->-- 



76 Varnishes of the Italian Violin-makers. 

sequent heating have been confined within reasonable 
limits, is perfectly soluble in oil of turpentine; 
generally, no unusual precautions need be taken in 
diluting the varnish with that vehicle. Oil is added 
to the vsimish to increase its cohesion and durability. 

"When the vehicle evaporates (after the application 
of the varnish) the pellicle is of a soft character. 
Wood, unless it is very dense, absorbs considerable 
quantities of these varnishes ; consequently the pellicle 
is adhesive. Such varnishes harden very slowly ; 
long after the surface has become so dry as to bear 
polishing, the varnish is sufl&ciently plastic to receive 
impressions when contact, accompanied by pressure, 
is long continued. Instruments covered with them 
require to be exposed to warm, dry air for many 
months before they are in a condition to bear handling 
and wear without injury. As may be supposed, the 
varnish remains perfectly elastic, and, as far as may 
be judged from the experience of some years, never 
loses that quality. 

In every respect, therefore, these yellow varnishes, 
with refractions of orange and light-red, are eminently 
adapted for covering musical instruments. To those 
who have a taste for delicate colouring, their iDeauty 
cannot be surpassed; but for others who prefer colours 
which are more vivid or more sombre (who find a 
lurid, stormy sunset more entrancing than that of 
perfect summer weather) it is necessary to provide 
varnishes of a more rufous character. Their mode of 
production will be now described. 

Having produced an oxy-terpene or resin in the 
way described, it is necessary to expose this substance 
to a careftiUy-regulated temperature at or above that 



OiUvamish from Turpentine Derivatives. 77 

of the water-bath, for certain specified times. The 
increased depth, as well as the particular shade, of 
colour will depend upon these factors — temperature 
and duration. 

It has been already explained that these resinous 
substances are singularly bad conductors of heat. 
They are at the same time very viscid ; conduction 
and convection are both wanting. To increase these 
attributes, it is convenient to add now to the oxy- 
terpene (which is supposed to remain in the basin in 
which it was formed) the quantity of pure siccative 
linseed-oil required for the constitution of the varnish : 
equal, generally, to from 30 to 50 per cent, of the 
weight of the initial colophony or Venice turpentine. 

The porcelain basin is removed from the water- 
bath to a sand-bath (or any other apparatus by which 
a gentle heat may be communicated) ; the tempera- 
ture is gradually raised above 100° C, while the con- 
tents of the basin are well stirred. In the course of 
an hour (if the rise of temperature has been slow) the 
colour of the varnish will be seen to be deepening ; 
this change will be increased and intensified, accord- 
ing to the duration of time and the degree of tem- 
perature, until a deep chocolate brown will be 
reached. The varnish may now be diluted with 
oil of turpentine, but this operation requires to be 
performed with caution. It is preferable to warm 
the essence to a degree nearly approaching the tem- 
perature of the varnish; if this precaution is not 
taken, the cold essence must be incorporated (by 
stirring) very gradually, the temperature of the 
varnish being meantime well maintained. 

If a still darker varnish is required, the heating 



78 Varnishes of the Italian Violin-makers. 

may be continued for a longer time, or the tempera- 
ture may be raised. The varnish, in bulk, will then 
after a certain time appear almost black ; when seen 
in a thick film on the surface of the white basin or bf 
the porcelain spatula, it will have a deep red tint ; 
in a thin film the colour remains a shade of yellow. 
To dilute this dark varnish requires "some practice 
and skill. If too much oil of turpentine be added, 
if too much be added at one time, if by dilution the 
varnish be cooled below a certain point (all factors 
incapable of exact definition), precipitation takes 
place, the varnish is spoiled. It must be made over 
again with new materials. 

Varnishes which have been heated in this way, so 
as to be subjected, wholly or in part, to a temperature 
appreciably exceeding ioo° C, have always a brown 
tint. 

It will be expected that the implements and tools 
used by the earliest violin-makers of Brescia and 
Cremona were a little primitive ; it is not to be sup- 
posed that a "balneum mariae" was to be found 
among them, although there is little doubt that it 
was known in the middle of the sixteenth century. 
At first it may be supposed that varnishes were 
made in a pot over a charcoal brazier or some such 
crude heating apparatus ; careless makers very likely 
reverted to this primitive apparatus long after the 
"balneum marisB" had come into general use. To 
some method of direct heating the examples of brown 
varnish, so often seen on many of the older instru- 
ments, must, without any doubt, be attributed. 

The uncertainty of result attending the use of the 
sand-bath, or any method of direct heating, soon 



Oil'Vamish from Turpentine Derivatives. 79 

induced me to abandon it in favour of the water- 
oven, whose temperature can be constantly main- 
tained at about 97° C. (the water-bath has also been 
occasionally used). Although the process is rather 
tedious (lasting from three to seven hours), the results 
are good in every way. Varnishes made by this 
means have not often been found difficult to dilute 
at the completion of the process. The requisite 
quantity of oil of turpentine, in a suitable flask, can 
be introduced into the water-oven previous to its use, 
so that it may acquire the same temperature as the 
varnish before admixture. 

Varnishes of pure shades of orange and red may 
be thus produced. 

Whatever the source of heat employed, it may be 
observed that when the solution (of oxy-terpene in 
linseed-oil) has been heated for some time at about 
100° C. or upwards, it begins again to froth; when 
stirred, bubbles of vapour rise through it. This 
vapour is steam. In every process employed it has 
been found that when an oxy-terpene is heated, as 
the colour deepens, minute quantities of water are 
given off. This water is not attributed to the residue 
of the nitric acid, but to chemical change in the 
terpene oxide, which becomes dehydrated (HgO is 
eliminated). 

As a consequence of this dehydration, the terpene 
molecule becomes (centesimally) more rich in carbon, 
losing two atoms of hydrogen and one of oxygen; the 
varnish, of which the resin is so important a con- 
stituent, becomes slightly less transparent; its ap- 
parent colour passes from yellow to orange, red or 
brown; its other physical properties change also — 



80 Varnishes of the Italian Violin-makers. 

it is less soluble in oil of turpentine, its viscosity is 
greater; when the pellicle of varnish is eventually 
formed it appears to solidify more quickly, and is of 
a somewhat harder quality than that formed from the 
oxy-terpenes. 

If the information given by M. Naudin (already 
quoted) is carefully considered, it will be evident that 
in the preparation of copal varnish the phenomena are 
identical with those which have been described. 
P3rrocopal is evidently a dehydrated oxy-terpene 
derived from copal. It appears that the qualities 
attributed to varnishes derived from copals of dif- 
ferent degrees of hardness are quite as likely to be 
due to the different temperatures to which the copals 
are subjected as to any material difference in the 
primary substances themselves. 

It will now be understood what an immense 
variety of varnishes, differing in apparent colour 
and in other respects, can be produced from the 
same original materials by modifications of the two 
factors of oxidation and dehydration. The propor- 
tion of nitric acid may be increased from lo per cent, 
to 30 per cent, and upwards ; the time allowed for 
nitrification may be a few hours or many days ; the 
extent of dehydration (which is obviously dependent in 
a great measure on that of oxidation) is limited only 
by the question of solubility : excess of oxidation and 
dehydration means absolute insolubility as far as an 
oil-v2imish is concerned. In general terms, the pro- 
gress of degradation (oxidation and dehydration) of a 
terpene hydrocarbon is marked by gradual decrease 
of solubility in oil of turpentine and linseed-oil and a 
similar increase of solubility in aqueous alcohol. 



(Hl-vamish from Turpentine Derivatives. 81 

In consequence of this it will be found that while 
with an oxy-terpene 30 per cent, of linseed-oil is 
quite sufficient to constitute a good varnish, 50 per 
cent, will be absolutely required for ^ solution as 
dehydration extends, with a still further increase as 
degradation proceeds. 

Given imperfect methods of determining the pro- 
portion of oxidant to terpene (no very easy matter 
where a fluid is concerned, for even now. the part 
capable of conversion to a resin can be only roughly 
estimated), a charcoal stove or some equally crude 
mode of heating, it would be practically impossible to 
produce two varnishes precisely alike. 

Another variable factor in the production of an 
oil-varnish is the proportion of oil, which influences 
not the colour only; but other physical properties as 
well. As to oxidation and dehydration, the desired 
colour of the varnish is some guide, experience 
enables the operator to produce a close approximation 
to a gfiven varnish ; but with respect to the proportion 
of oil to resin, so far no rule has been found. Practical 
copal-varnish makers appear to test a drop of the 
solution of oil and resin, cooled on a plate of glass, 
judging by its character as to the sufficiency of the 
oil. With a hard resin this may be practicable, but, 
with one that is comparatively soft:, no reliable 
indication is apparent. 

The quality of a varnish, apart from its colour and 
brilliancy, can be judged only after it has become 
hard enough to be finished and polished. Under 
ordinary circumstances, in this climate, this is a 
question of at least six months. Experience must be 
the outcome of many years. 

V. G 



82 Varnishes of the Italian Violin-makers. 

The boiled linseed-oil of commerce, even the better 
qualities used for artistic purposes, will be seldom 
found of sufficient purity for fine varnish. In the 
early stages of these experiments, great difficulty was 
experienced in obtaining a supply of siccative oil fi-ee 
firom mineral impurity. It is more than probable that 
the existence of equal difficulty in the olden times led 
eventually to the invention of the Neapolitan method, 
in which raw linseed-oil may be directly used. But 
in avoiding Scylla the old varnish-makers ran great 
risk of falling into Charybdis, as a little consideration 
will show. 

The main object of the introduction of oil into a 
varnish is to increase its cohesion and toughness. 
The extent of the oxidation of a drying-oil to render 
it suitable for this purpose is small : from 5 to lo per 
cent, of nitric acid would be amply sufficient. If 
oxidation is carried much further than this the oil 
becomes resinous, its essential attribute is destroyed. 
Neapolitan varnishes have been made (in the course 
of these experiments) which, when dry, pulverized by 
friction precisely in the same way as turpentine- 
varnishes (without oil). No doubt many such var- 
nishes were made and used in the olden time ; as a 
consequence, some modern experts have designated 
them as spirit-varnishes, in the same way as F6tis so 
described the varnishes of Maggini. 

The Neapolitan varnish-maker was consequently 
on the horns' of a dilemma : on the one hand he was 
restrained by the consideration that oxidation beyond 
a certain point destroys the most valuable property 
of his oil (as far as an oil-varnish is concerned) ; while, 
on the other hand, the showy red varnish, which he 



Oil'Varnish front Turpentine Derivatives. 83 

often desired, could only be reached through oxidation 
far beyond this point. The oil was required to play 
a dual role, which required the development of two 
totally different attributes. 

The manufacture of yellow varnish was easy 
enough, for in this case a solution of colophony or 
Venice turpentine in raw linseed-oil could be nitrified 
and oxidized with a proportion of nitric acid com- 
patible with the retention of the usual properties of a 
drying oil (say firom 5 to 15 per cent, of the combined 
weight of the terpene and linseed-oil). Such varnish 
is often of excellent quality, and, if well made, not 
essentially different fi-om a Cremonese or Venetian 
varnish made in the usual way — always understood 
that the oxidation of the oil is apt to be as much in 
excess as that of the terpene is too little. The pale 
varnishes of Alessandro Gagliano, of Naples, are 
examples of this class ; M. Vidal writes of his instru- 
ments : — " His workmanship is careftil and skilftil ; 
" his varnish only leaves something to be desired, 
notably in its nuance of a yellow greyish tone ( jaune 
fond grisitre), little pleasing to the eye." The oxida- 
tion of the terpenes is manifestly too little. 

But for a red varnish, the transformation of part 
of the oil to a red-brown resin and the retention of the 
remainder in the form of a boiled oil are obviously 
incompatible with simultaneous treatment. The best 
mode of compromising this difficulty appears to be 
either : ist. To add to the terpenes (colophony and/or 
Venice turpentine) a small amount of linseed-oil (raw), 
say about 10 per cent, of their weight ; this oleo- 
terpene is diluted with oil of turpentine to a convenient 
consistency, nitrified with 25 or 30 per cent, of nitric 

g2 



u 



84 Varnishes of the Italian Violin-makers. 

acid, decomposed, dehydrated in a water-oven to the 
required depth of colour (4 to 6 hours) ; the resulting 
oleo-resin is dissolved in boiled linseed-oil (produced 
either in the ordinary way or by the oxidation of raw 
oil with a small proportion, such as 5 per cent., of 
nitric acid), as though it were a simple terpene resin, 
finally diluted with oil of turpentine as customary. Or, 
2nd. In nitrifying the oleo-terpene as just described, 
about 5 per cent, less of nitric acid is used ; after de- 
composition and dehydration, the oleo-resin is dissolved 
in raw linseed-oil and a further oxidation of the varnish, 
with 2\ or 5 per cent, of nitric acid, is effected, again 
followed by dilution like an ordinary varnish. Of 
these two alternatives, in my hands, the first has 
given the most satisfactory results. Tononi, of 
Bologfna, has left excellent examples of varnish of 
this kind. 

It is somewhat diflficult to imagine violin-makers 
like Gagliano and Tononi grappling with problems of 
this sort ; here again the assistance of some student 
of chemistry appears probable. 

As may be supposed, pale Neapolitan varnishes 
are perfectly transparent ; the decrease of this quality 
in the red varnishes of this t3rpe is very apparent on. 
the tables of the instruments which they cover ; the 
backs and ribs are effective — ^there the slight want of 
transparency does not attract attention. 

There is no difference whatever in the preparation 
of a varnish whether the initial ingredient is colo- 
phony, Venice turpentine, or oil of turpentine. The 
whole of the colophony will, however, remain in the 
varnish, whereas in the other cases an unknown 
quantity, being volatile and incapable of conversion 



Oil-varnish from Turpentine Derivatives. 85 

to a resin, will be sooner or later evaporated and lost. 
In estimating the proportions of oxidant and oil, about 
80 parts of colophony have been looked upon as 
equivalent to 100 parts of Venice turpentine and to 
about 160 parts of oil of turpentine. 

Colophony and Venice turpentine do not produce 
varnishes of quite the same colour or quality, whether 
they are. used in making Cremonese or Venetian 
varnish, or form part of a Neapolitan varnish. The 
larch product produces a varnish which is slightly, 
but appreciably, more pure in colour and more resis- 
tant than that which contains colophony. As far as 
experience goes, an instrument treated with Venetian 
varnish gives a tone which is more brilliant than that 
resulting from the softer Cremonese — as might be 
expected from the superior elasticity of pine wood as 
compared with that of larch. 

Tone results lying between those attributed to 
these two types of varnish can be produced by 
blending Venice turpentine either with colophony or 
with oil of turpentine (in the latter case the nitrifica- 
tion process is prolonged) ; such mixtures of natural 
terpenes give rise to colours of a warmer tint than is 
obtainable by the use of either colophony or Venice 
turpentine alone. 

« * » « • 

On reading the descriptions of the old varnishes 
by various writers, it is significant that, while com- 
paratively pale varnishes are, without exception, 
described as lightly and evenly laid on, those of a 
deeper tone are very often represented as having 
been thickly applied, crusted, used with want of care 
and skill, or as being more or less clotted (that these 



86 VamisAes of the Italian Violinrtnakers. 

descriptions are true I can myself aflfirm). If the 
varnishes had been formed (as surmised) from a 
common basis, coloured to suit the taste of individual 
violin-makers, it would not be easy to explain this 
general agfreement of description. He must be a 
clumsy operator indeed that cannot succeed in pro- 
ducing, from a given uniform basis varnish, coloured 
varieties approximating in qualities affecting their 
application. 

When, however, we turn from these coloured con- 
coctions to the pure oil-varnishes, whose mode of 
preparation has been described, an explanation is at 
once apparent : it must be looked for, not in the want 
of care or skill on the part of the artisan, but in the 
inherent properties of the varnish itself. A yellow 
varnish, whose basis is an oxy-terpene, is limpid, 
(with a little practice) easy of even application by 
means of a brush ; but when this same varnish is 
heated for a few hours over a water-bath or in a 
water-oven, or for a much shorter time over a sand- 
bath, it becomes viscous, requiring a great deal of 
skill with the brush to lay on passably well. Neither 
can this deep-coloured, dehydrated varnish be diluted 
with oil of turpentine sufficiently to remove this 
practical inconvenience. It can be done by a large 
increase of oil, but then the effect of such a pro- 
ceeding on the tone of the instrument, as well as on 
the brilliancy and colour of the varnish, must be con- 
sidered. My opinion is that the old makers were 
quite right in rejecting this remedy ; there is no room 
for doubt that those who take the trouble to study the 
question will unanimously endorse the conclusion at 
which I have arrived without any hesitation. 



Oil'Varnish from Turpentine Derivatives^ 87 

That there should be no possilpility of doubt on 
this important question, the same varnish was made 
several times (involving a large expenditure of time 
and labour), in order to determine with exactness the 
minimum proportion of oil and limit of dilution with 
oil of turpentine consistent with its constitution. 
The details of this experiment will now be given, 
since the information which it affords and the light 
which it throws on the proceedings of the old makers 
are both instructive and interesting. 

One hundred parts (1,200 grains or 80 gj-ammes) of 
colophony were dissolved in 50 parts of oil of turpen- 
tine ; this viscid fluid was nitrified with 33 parts of 
nitric acid (1*420), the nitrification being slow and 
extending over five or six days. The nitro-terpene 
was a dark-brown, plastic substance, almost solid ; it 
could not be stirred with a porcelain spatula, but was 
cut out of the deep porcelain basin (6J inches or 
16 centimetres diameter) in which it was prepared, 
with a knife, in suitable pieces for decomposition in a 
similar basin. 

Decomposition was effected over a water-bath in 
the usual way. After the reaction, the resin was 
heated in a water-oven (with firequent stirring to 
facilitate the elimination of the aqueous residue of the 
nitric acid) for about an hour. Result — a yellow 
resin of a brilliant orange colour when seen in bulk. 
Dissolved over the water-bath, in 50 parts of linseed- 
oil (boiled), then diluted with 150 parts of oil of 
turpentine. The same varnish was made again and 
diluted with 180 parts of oil of turpentine. Both 
these varnishes were brilliant pale varnishes of a pure 
primrose-yellow colour when thinly spread on a white 



88 Varnishes of the Italian Violin-makers. 

surface ; a drop was copper-coloured. The red rays 
from a reading-lamp could easily pass through a 
thickness (in a flask) of about 3 inches diameter. 

The oxy-terpene was made again (under identical 
conditions], but the resin was exposed to the heat of a 
water-oven (about 97® C.) for six hours (instead of one 
hour) ; it then became of a deep red colour when seen 
in bulk. It was dissolved in 50 parts of boiled 
linseed-oil and (being very thick and viscous) it was 
diluted with 250 parts of oil of turpentine. Hot, the 
varnish appeared to be good, but on cooling it became 
turbid ; precipitation was so great that the varnish 
was useless. 

The varnish was repeated again, the proportion of 
linseed-oil was increased to 55 parts, the oil of tur- 
pentine reduced to 180 parts. Precipitation, on 
cooling, was again considerable ; after subsiding for 
a day or two, the decanted varnish was bright and 
fairly good. 

Again the varnish was made — ^the linseed-oil 
increased to 60 parts, oil of turpentine diminished to 
150 parts. The result was a perfectly bright, clear 
varnish, which required neither to be decanted nor 
filtered. 

A thin film of this varnish is still yellow, but as 
the film thickens the colour deepens to a deep orange- 
red. A drop is the colour of a ruby. It is much less 
translucent than the first varnishes ; the rays of the 
reading-lamp can no longer pass through a flask of 
it, although it has not the faintest trace of turbidity. 

The viscosity of this varnish is very much greater 
than that of the first-made yellow varnishes; the 
latter are perfect varnishes for facility of application, 



Oil-varnish from Turpentine Derivatives. 89 

but the former is difficult to lay on evenly without 
clots. 

The 250 parts of oil of turpentine which were 
originally designed for this varnish would not have 
been too much to bring it to the same thickness as the 
yellow varnishes; but, as was clearly proved, even 
180 parts were incompatible with the constitution of 
the varnish, with the assistance of an increase of 
5 parts in the oil. 

Were the resin heated still longer in the water- 
oven, it would become still more viscous and still less 
soluble in both linseed oil and in oil of turpentine. 
Dissolved in the same proportion of linseed-oil (60 
parts), the dilution with oil of turpentine would 
have to be diminished, increase being manifestly 
impossible ; the varnish would be of a deeper colour, 
but the difficulties of application would be augmented. 

In the case of these deep-coloured varnishes, made 
from dehydrated oxy-terpenes, the conditions with 
which the operator has to contend are precisely 
similar to those attending the production of varnish 
from hard copal. It requires both experience and 
judgment to rightly determine the proper proportions 
of oil and essence. After the varnish has been com- 
pleted and cooled, if it is too thin or too thick there is 
no known remedy ; it cannot be heated again with a 
view to evaporation or dilution ; if any attempt (ever 
so carefully performed) be made in this direction, the 
varnish will become turbid on cooling and will be 
spoiled. If, however, a varnish has been made which, 
on cooling, is found to be too thick or too thin, 
another similar varnish can be made with an equal 
tendency in the opposite direction ; when quite cool. 



90 Varnishes of the Italian Violin-makers, 

these two varnishes may generaUy be mixed in aU 
proportions. 

With all the varnishes, one element which militates 
against their perfectly even application is their ab- 
sorption by the wood ; if this absorption were uniform 
it could not affect the question, but its extent varies 
with the grain as well as the direction of the section 
of the wood. Naturally, when the section is parallel 
to the longer axes of the cylindrical fibres, the absorp- 
tion of varnish is far less than when it is at an angle 
to these axes. Owing to the moulded shape of the 
back and table of violins, as well as to the variations 
in the grain of the wood, the angle of section is never 
constant. With every coat of varnish, therefore, 
especially the earlier coats, uneven absorption occurs, 
of necessity producing to some extent an uneven 
surface. Whether this was remedied by the old 
makers by "rubbing down" before the last coats 
were put on, or whether the varnishing was com- 
pleted and then rubbed down and polished, can 
only be surmised : probably individuals had different 
methods in this respect. 

No mode of reducing the finished varnish to a 
good and true surface has been found to equal rubbing 
down with a rubber kept well moistened with alcohol 
of 87 to 90 per cent. Anyone who is expert at French 
polishing can perform this operation with perfect 
success without removing the smallest fraction of the 
varnish. It would be interesting to know whether 
Stradivari knew of this process. 

Reference has already been made to the sugges- 
tion of M. Mailand that in order to prevent absorp- 
tion of the varnish by the wood, it should be previously 



Oil-varnish from Turpentiru Derivatives. 91 

sized. This suggestion appears to arise from a total 
misconception of* the whole varnish question ; it needs 
no further discussion. 

The various colours of the varnishes depend not 
on their actual colour (that which is seen when they 
are spread in a thin film on a smooth, white surface), 
but on that which is apparent when the thickness of 
the pellicle is sufficiently increased to allow of com- 
plicated refraction and reflection from varying planes. 
If their colour were a substantive one (the simple 
absorption of rays situated in one part of the spec- 
trum and the reflection of the complement), such as 
that of a pigment, the different varnishes would cause 
little variety ; the tone would be always yellow, tend- 
ing more or less towards brown. 

It will therefore be understood that the effect pro- 
duced will mainly depend upon the quality of the 
varnish, the thickness of the film, and lastly, not 
least, the character of the wood on which it is laid. 

When a new varnish is made it may be tested by ap- 
plication to a strip of maple or other wood ; the result 
is some guide as to the probable appearance of a violin 
when covered with it, but to a very limited extent. 
Two backs of maple wood (both of fine figure and 
quality, but differing in the angle of section or in the 
original character of the wood) may be treated with 
the same varnish under precisely identical conditions; 
when finished it is difficult to believe that the varnish 
on the two instruments can be the same, so different 
is their apparent colour and general effect. 

Apart, therefore, from the great variety in the 
characters of the varnishes, the additional factors of 
thickness of pellicle and of variation in the section or 



92 Varnishes of the Italian Violin-makers. 

nature of the wood add to the complexity of the final 
results. It seems to be quite impossible to predict, 
with any precision, what will be the effect of the 
application of a known varnish to a virgin instru- 
ment. 

There can be no doubt that some of the old violin- 
makers used repeatedly wood fi-om the same log or 
tree; in this way the wood factor would be nearly 
a constant (not entirely so, for in the same tree 
adjoining sections will vary to a considerable extent). 
Keeping this one factor of wood an approximate con- 
stant, they may have striven, by change of varnish, 
towards their ideal of perfection. It is not at all 
difficult to understand that this pursuit of an ideal 
was fascinating ; while the uncertain and surprising 
variety of results precluded their occupation from 
becoming monotonous. 

In an Appendix will be given the details of a 
number of varnishes which have been prepared and 
used for musical instruments as well as for other 
purposes. 



/ 



( 93 ) 



Chapter VI . 

INTERESTENG CHEMICAJL PROBLEMS WHICH HAVE 
PRESENTED THEMSELVES DURING THE PROGRESS 
OF THIS RESEARCH. 




►HE earliest observation of dichroism in terpene 
products which has come under my notice 
is that of M. Berthelot, in 1853. It is to be regretted 
that he paid (or seemed to pay) no attention to it ; at 
any rate, he unfortunately did not suggest any 
explanation of the phenomenon, which is to be under- 
stood in this restricted sense, that whereas the colour 
of transmission is yellow, that of refraction is red. 
Only once divergence from this general rule has been 
observed — an alcoholic solution of terpene derivatives 
from the bark of picea was seen to be yellow in a 
thin film, while in bulk the colour was purple. Neither, 
as far as I know, is any explanation to be found 
elsewhere, which is the more extraordinary, as it 
seems to be characteristic of a certain class of organic 
substances and its true significance cannot be without 
importance. 

A striking instance of its occurrence is to be seen 
in the following experiment : — If oil of turpentine 
(rfpinene), well rectified and colourless (sp. gr. 0*869 » 
(a) y = + 62*5 at 18° C), is mixed with aqueous 
alcohol (sp. gr. 0*830 at 18°), about five parts alcohol 
dissolve one part of pinene (by volume). To this 
alcoholic solution 5 per cent of nitric acid (1*420) is 
added ; a large globule of oil precipitates. The next 



I 



94 Varnishes of the Italian yio/in^makers. 



day a further 5 per cent, of nitric acid is added, more 
oil separates ; the whole of the separated oil (which 
amounts in volume to about one-fourth of that origin- 
ally in solution) now rises to the surface. Shaken 
together at intervals, in four days the oil completely 
re-dissolves, solution again becomes perfect, all tur- 
bidity disappears. On the fifth day a fiirther 5 per 
cent, of nitric acid is added, the same on the sixth day, 
without causing any change or turbidity (the alcohol- 
pinene solution now contains 20 per cent, of its volume 
of nitric acid). The solution slowly approaches a pale 
brown colour. In ten days, from the first addition of 
nitric acid, the colour has become distinctly yellow- 
brown (sherry), without any sign of dichroism. After 
standing for a month, in difiused daylight at about 
1 6° C, the solution has become strongly dichroic, 
appearing in bulk of a deep red colour ; it resembles 
a solution of oxidized colophony standing beside it, 
but is the darker of the two. A considerable amount 
of nitrous ether has been formed, the yellow-brown 
colour evidently denoted oxidation, pure and simple, 
of the pinene. My interpretation of the dichroism is 
that, with increased oxidation, the pinene has split ; 
the solution now contains two terpenes having 
different optical properties : the red colour is due to 
interference. 

The flask containing this solution is closed by a 
caoutchouc cork provided with a small glass tube 
drawn to a capillary point. Curiously enough, quite 
contrary to my expectations, this flask and its contents 
gradually — slightly but appreciably — increases in 
weight. After standing for several months, in strong 
difiEiised light, the colours are unchanged. 



Interesting Chemical Problems. 96 

Another experiment :— lo grammes of colophony 
are slowly dissolved (at about i6°) in lo c.c. of alcohol 
(o*83o), one part of the former is therefore soluble in one 
of the latter at i6°. To this solution is added 5 c.c. of 20 
per cent, nitric alcohol (i to 4) ; turbidity immediately 
ensues ; in half-an-hour the solution divides into two 
(about equal) strata, the upper more pale than the 
lower. These two strata are frequently well shaken 
and mixed together at intervals. The next day a 
mass of nodular crystals have separated ( ? sylvic 
acid). The crystals are filtered from the fluid, which 
is of a yellow-brown colour, strongly dichroic (it 
evidently still contains some of the crystallizable 
substance in solution). The crystals are much less 
soluble in the alcohol than the initial colophony, at 
least five times less ; their solution is " sherry " coloured, 
not dichroic. When fiirther oxidized by nitric alcohol, 
in a month it becomes (in diffused light) slightly 
dichroic, a thin film is pale yellow tinged with brown. 
This flask and its contents, closed also with a caout- 
chouc cork and capillary-pointed tube, slowly gains 
weight. The solution firom which the crystals were 
separated, also further oxidized in the same way in a 
precisely similar flask, slowly loses weight. 

The preparation of nitro-pinene (dextro oil of 
turpentine plus 30 per cent, of nitric acid) has already 
been described in detail. To effect decomposition, the 
brown fluid, separated firom the aqueous residue of the 
nitric acid, is placed in a capacious porcelain basin 
over a boiling water-bath. As soon as the temperature 
of the nitro-pinene, in contact with the surface of the 
basin, approaches that of boiling-water, a reaction 
commences at the margin and soon extends, fiimes of 



96 Varnishes of the Italian Vtolin-fpuikers. 

the lower oxides of nitrogen are thrown off, accom- 
panied by steam ; the reaction becomes general and 
rather violent, petty explosions eject small particles 
from the contents of the basin (these explosions are 
doubtless due to steam, produced by the heat of the 
reaction, which is prevented from escaping freely by 
the terpene). When the reaction is complete, the 
unpromising dull brown fluid has become converted 
into a brilliant yellow varnish, which, in bulk, appears 
of a rosy red colour. From this hot varnish clouds of 
vapour are escaping. The odour has again changed 
from that of fir- or pine-wood to a pleasant aromatic 
one. All water has entirely disappeared; the contents 
of the basin, on cooling, are found to be perfectly 
homogeneous, bright, clear, free from turbidity. This 
viscid, oily varnish is very slightly soluble in the oil 
of turpentine from which it was derived, when cold ; 
warmed with it, it dissolves completely, but the bulk 
of the resin is precipitated on cooling. The varnish 
consists of a yellow resin (rosy red by refraction), 
dissolved in a volatile oil. 

No varnish that has been made has had an odour, 
when cool, of oil of turpentine. The odour of this 
substance is so characteristic that it is difl&cult to 
believe that it could escape detection even though 
accompanied by aromatic substances. The volatile 
oil which takes its place appears to agree closely with 
it as to boiling point (about 155° to 156°, but not 
constant); specific gravity, 0*86 1 ; optically inactive. 
This agrees with terebene, with which its ethereal, 
faintly aromatic odour also corresponds. As may be 
understood, I had seldom opportunities for examining 
this volatile oil, but when I did so I found it in- 



Interesting Chemical Problems. 97 

variably free from optical activity ; this circumstance 
caused me some surprise, it appeared to me possible 
that, in some way, its activity had been destroyed ; an 
attempt was made to determine this question. As 
the experiment is deemed to be of some importance, 
its details must be given. 

Into a capacious tubulated glass retort were poured 
loo c.c. of well-rectified oil of turpentine (supposed to 
be American) which was perfectly limpid and colour- 
less. Its sp. gr. at 17^° was 0*869. In a 254 m.m. 
tube it gave for the transition tint, at 20°+ 138° (my 
thirty years old Dubosc instrument and my old- 
fashioned methods will be excused). 

With these loo c.c. of pinenewere carefully mixed, 
drop by drop, 5 c.c. of nitric acid. At intervals, during 
four days, the oil and the residue of the nitric acid 
were thoroughly mixed together. On the fifth day, 
the nitro-pinene was of a golden yellow colour, the 
residue of the nitric acid of a similar colour, a shade 
darker ; the latter was now withdrawn by means of a 
pipette, some fragments of glass were introduced into 
the retort and it was placed over the water-bath. 

When the water in the bath began to boil, bubbles 
of nitrogen, without colour or odour, rose through the 
fluid ; the nitric acid was, apparently, completely 
reduced. When the temperature of the fluid reached 
85° the reaction was fairly brisk ; the evolution of gas 
reached its maximum at about 100° C. On signs of 
subsidence of reaction a sand-bath was substituted 
for the water-bath. At this stage the oxy-pinene 
was of a clear amber-yellow colour, no indication of 
dichroism. With the sand-bath (a gentle heat being 

V. H 



98 Varnishes of the Italian Vtolin-makers. 

applied) the temperature soon rose to 105°, at which 
a certain amount of reaction was still apparent ; at 
110° the evolution of gas practically ceased. At this 
temperature the colour of the liquid (hitherto constant) 
suddenly changed to a fine rosy red (dichroism) ; soon 
after it became txirbid, then brown, finally (during 
the advanced stages of distillation) almost black. 

At 145° the terpene began to boil (both oil and 
water condensed) ; it boiled briskly at 150° ; tempera- 
ture gradually rose to 153°, the liquid boiling very 
briskly. The oil now coming over had a pale gfreen 
fluorescence. Dilute aqueous sodic carbonate (cold) 
dissolved fi"om this a little oil of a yellow colour. The 
oil was well washed with distilled water, then dried 
with pure calcium chloride, when it was quite colour- 
less : — sp. gr. 0*863 ^^ 19°; rotation in 254 m.m. tube 
at i8°— 64°. The fluid part of the contents of the 
retort was now poured into another retort ; the solid 
part was left behind. On resuming distillation, tem- 
perature rose to 158° ; the distillate, washed and dried 
as before, was not quite colourless (not so easy to 
read in the polarimeter) : — sp. gr. o*86o at 19° ; rotation 
in 254 m.m. tube at i8°— 67/68°. 

The first part of the distillate contained perhaps a 
little dextro-terpene. 

Ignoring temperature, .we get (for the formula 

(a) / = -j-t) for the original pinene (a) / = + 62*5° ; 

for the distilled derivative (the mean of ^67/68°) 
(a) j = - 30-9°. 

The rotation of the inverted terpene approaches 
very closely to one-half of that of the original fluid. 

Except that it was Isevo-gyrose, this volatile oil 



Interesting Chemical Problems. 99 

could not be distinguished from that which I have 
always looked upon as terebene. 

The dark brown substance remaining in the retorts 
was not a resin, but a soft aromatic substance. If, 
however, oxidation is carried ftirther, as has been 
shewn with 30 per cent, of nitric acid, pinene is 
partially converted into a resin of fine quality. So 
far, my efforts to obtain a resin from the volatile oil 
(or terebene), which is the other product of the oxida- 
tion of pinene, have not been successftil. 

Professor William Allen Miller gives a definite 
theory of the formation of resins (Organic Chemistry, 
1869). He writes (625) : " In the majority of cases the 
" resins are formed by the oxidation of the essential 
" oils contained in the trees that yield them ; hence it 
" is not surprising that in many instances they have 
the composition of oxides of the hydrocarbon C20 
H32, or of a hydrocarbon derived from this, which 
" has lost a certain number of atoms of hydrogen in 
" exchange for half that number of atoms of oxygen." 
My efforts to discover a more modem theory, ex- 
pressed in definite terms, have not been successfiil. 

If oil of turpentine {d pinene) may be taken as the 
t3rpe of the essential oils, it is certain that it may be 
oxidized (without other apparent change), by means 
of nitric acid, in the way described. Whether poly- 
merisation accompanies this oxidation I do not knoW ; 
perhaps it does. If oxy-pinene is subjected to heat it 
appears to split; the homogeneous fluid exhibits 
dichroism, quickly followed by turbidity. If oxida- 
tion has been considerable, and if the heat does not 
exceed 100° C, only dichroism is observed — solution 

h2 



it 
it 



100 Varnishes of the Italian Violin-makers. 

remains perfect. A volatile hydrocarbon is evaporated 
or distilled, which under different conditions may be 
either optically inactive or laevo-gyrose. This hydro- 
carbon resembles the original pinene in density and 
boiling-point, but is quite different from it in its 
behaviour with nitric acid and its affinity for oxygen, 
which is far less. After many experiments and ob- 
servations, the conclusion arrived at is that, let 
oxidation be much or little, drastic or very gentle, 
only a part of the terpene hydrocarbon is capable of 
conversion into a resin or resins. If this important 
fact should be confirmed by future investigators, the 
very interesting question must arise as to what 
happens, in Nature, to the other part of the hydro- 
carbon which is not transformed into a resin. 

With respect to the oxide or resin, it may be 
dehydrated at ioo° without decomposition or without 
becoming insoluble in its accompanying hydrocarbon ; 
it may then be oxidized again, and again dehydrated. 
The limit of repetition of this double process of de- 
gradation has not been reached ; difficulties of 
manipulation arise which, for an amateur, are not 
easy to deal with. 

It does not appear, by any means, that dehydra- 
tion is a necessary result of oxidation : its extent is 
determined by conditions which promote the elimina- 
tion of water. It is essential for dehydration, however, 
that there shall be sufficient oxygen available — ist, 
to combine with two atoms of hydrogen ; 2nd, to 
saturate an atom of carbon from which two atoms of 
hydrogen are detached ; in consequence, two atoms of 
oxygen are necessary for every two atoms of hydrogen 
lost by the hydrocarbon. Experiments have shown 



Interesting Chemical Problems. 101 



that the extent of oxidation influences that of the 
succeeding dehydration. 

The effect on solubility of this degradation of the 
hydrocarbon is not difiicult to follow: increase of 
degradation causes decrease of solubility, firstly in oil 
of turpentine, secondly in linseed-oil ; as solubility in 
these two menstrua decreases so does it steadily 
increase in aqueous alcohol. Affinity for oxygen 
increases as degradation proceeds. 

Temperature not exceeding ioo° appears to have 
no effect on an oxide or a resin beyond dehydration, 
but when higher temperatures are resorted to very 
complicated effects of decomposition ensue. These 
phenomena have been carefully studied and are well 
described by Rabat6 and others. The brown colours 
produced whenever a sand-bath or other mode of 
direct heating is used are obviously due to one or 
more of these decompositions. 

« . * « « * 

It has been surmised that with respect to varnish 
(or to a plastic resin) drying and solidification are 
not the same thing ; for I have observed that all the 
varnishes continue slowly to solidify or harden for 
years after the laying on has been completed, while I 
am aware that, under the favourable conditions to 
which they are exposed (fairly dry air of a uniform 
temperature of about i6°), they must cease to lose 
weight by evaporation in a comparatively short time. 
It is impossible to demonstrate this with a violin 
because, as has been already shown, its own weight 
is never constant. A method of demonstration was 
arranged which is not open to this or, apparently, to 
any other objection. 



102 Varnishes of the Italian Violin-makers. 

Four slips of polished plate-glass (3 inches X i) 
were, on one side, each coated with a different varnish. 
The surface of the glass was carefully cleaned, the 
varnish was poured on to it, the excess being drained 
off at one comer. The exact weight of each glass 
plate was known, they were subsequently weighed at 
specified times. The composition of the four varnishes 
was as follows : — 

No. I. No. 2. No. 3. No. 4. 

Colophony - - . 23*80 25*64 

Venice turpentine - - 25*50 36*60 

Oil of turpentine (as solvent) i6*6o i8*oo 25*50 

Linseed-oil - - - 11*90 15*36 12*24 12*20 

Oil of turpentine (as diluent) 4770 41*00 36*76 51*20 

100*00 100*00 100*00 100*00 



The oil of turpentine used as solvent was employed 
to dissolve and thin the colophony or Venice turpentine 
previous to oxidation ; a considerable part of it may 
therefore be expected to have been oxidized, a further 
unknown quantity evaporated in the subsequent pro- 
cesses. The oil of turpentine used as diluent was 
stirred into the varnish at the last stage of its prepara- 
tion ; as this operation was performed as expeditiously 
as possible (at a temperature below 100°) and as the 
varnish, when complete, was immediately, in all cases, 
transferred firom the basin to a glass flask, then corked, 
the loss of this oil of turpentine cannot have been 
large. 

No. I, after oxidation, was heated in a water-oven about i hour. 
No. 2, „ „ „ „ „ „ „ 6 hours. 

No. 3> >> >> »» »» >» >> >> o ,jj 

No. 4) M »> >»" n » >» » 7 >» 

















1 

1 

1 


y- •• • '» •,■ :' 








Obsi 


■ 
- 








\ 


i ; ■ ■ 








. — 








Class and Varnish. 


Veight. 


' V 






. 


No.l. 


No. 2. 


No. 8. 


• 


No. 4. 




• • 

• • 




Weight of glass . . 


94*70 


12676 


133-38 




Date. 

1902 

3 Nov. . . 


Temp. 

cr 


98*80 


133*48 


136-70 




' 


4 »» •• 


28° 


9760 


13252 


136-10 




12*60 


t 


5 ,. •• 


26° 


97-50 


132*33 


136-00 




3*07 


\ 


6 „ .. 


23" 


97-46 


132235 


135*97 


/ 


1*43 




7 „ .. 


24° 


97*43 


132-17 


135*94 




1*12 


r 


8 „ .. 


28° 


97*40 


132-11 


135*92 




•82 




9 „ .. 


28° 


97*39 


132-07 


135*90 




•61 




17 „ .. 


24° 


97*36 


131*95 


135*85 




2-15 




24 „ .. 


28° 


97*36 


131-92 


135*835 




-72 




30 „ .. 


28° 


9738 


131 92 


135*83 




*4i 




7 Dec. . . 


21° 


97*39 


131-90 


135-81 




*4i 


1 


14 ,, .. 


26° 


97*40 


131*92 


135*815 




•00 




21 „ .. 


24° 


97*40 


131*93 


135*815 




•10 




28 „ .. 

1908 

iSJan. .. 


22° 
28° 


97.41 
97*42 


13^*933 
131-96 


135*818 
135*83 




•10 






23*54 




14 Feb. .. 


22° 


97*42 


132-00 


135*845 






12 April . . 


28° 


97*43 


132-02 


135*85 






2 June . . 


28° 


97*44 


13205 


135*87 




' 


II Oct. .. 


26° 


97*46 


132-07 


135*875 




1 


31 Dec. .. 

1904 

23 May . . 


22° 
22° 


97*47 
97*49 


132-07 
132-105 


135-878 
135-89 


i weight. 

) 2-01 








__ 







To /ace p. 102.] 



r 



1 



1 . 



\ 



\ 



/ 

Interesting Chemical Problems. 103 



On the 14th February, 1903, all the varnishes ceased 

o be adhesive, but they were all still comparatively 

joft. During the progress of the experiment there 

was no visible change of any kind in either of the 

varnishes. 

The increase of weight after the minimum has 
been reached doubtless denotes absorption of oxygen 
from the air (comparatively dry, with a constant tem- 
perature of about 25°) ; under the circumstances, 
hydration is highly improbable. 

It seems to be perfectly clear that part of the oil 
of turpentine, used as diluent, is permanently retained 
and forms part of the dry pellicle ; this appears to be 
beyond all doubt. If the increase of weight, after the 
minimum, is rightly attributed to absorption of 
oxygen, this seems, from the relative proportions, to 
be referable to such part of the oil of turpentine and 
not to the linseed-oil. 

This experiment clearly proves what has been 
surmised from numbers of less exact experiments — • 
that solidification of a varnish or of a resin is generally 
quite a different thing from drying (evaporation of the 
volatile part of the pellicle) ; oxidation appears to play 
some part in the question, but it may be that, in the 
main, solidification denotes a gradual modification of 
physical form, perhaps polsrmerism. 

If it is supposed that oxidation goes on after the 
evaporation has ceased, it must be equally assumed 
that a certain amount of oxygen is absorbed during 
the continuance of evaporation. In this case the 
figures recorded for the loss by evaporation must be a 
little too low. This consideration introduces com- 
plication into the problem, but cannot materially 



104 Varnishes of the Italian Violin-makers. 

influence the main conclusions to be drawn from the 

facts. 

* * • « * 

In the production of oil-varnish, whether the resin 
ingredient be natural or artificial, some interesting 
problems introduce themselves. These soon attracted 
my attention, and during the preparation of such 
varnishes I was constantly on the alert for any clue 
that might lead to their solution. Close observation 
added very little to the known facts, however ; but a 
little enlightenment on some of the factors involved 
now enables probable explanations to be suggested. 

The direct heat to which copal is subjected to pro- 
mote its solubility leads to very uncertain results; 
the evident object of the operator is to fuse the fossil 
resin without allowing much time for dehydration. 
Fusion alone is not what happens to the resin; it 
splits into a volatile hydrocarbon and a colophone; 
the solubility and brown colour of this latter substance 
depend upon the unknown extent of its dehydration. 

The case with the artificial resins is very much the 
same when a sand-bath or other direct-heating method 
is employed — ^brown colour results from decomposition 
and the extent of dehydration is an unknown quantity; 
but temperatures not exceeding ioo° C. appear to 
have no effect except to cause dehydration, while the 
extent of this may be approximately estimated from 
the degree of oxidation and the duration of a fairly 
constant temperature. 

Experience leads me to suppose that persistent 
degradation (substitution of one atom of oxygen for 
two of hydrog^ti) of a terpene hydrocarbon is attended 
by constantly decreasing solubility in oil of turpentine 



Interesting Chemical Problems. 106 

and linseed-oil ; by increasing solubility in aqueous 
alcohol, as well as affinity for oxygen. 

Oxy-terpenes, which have lost little hydrogen, are 
readily soluble in both oil of turpentine and linseed- 
oil ; consequently the proportions of the varnish are 
determined by considerations quite apart from solution ; 
but with loss of hydrogen they become less and less 
soluble in these . menstrua as they become more 
viscous (an indication of degradation) ; a point is soon 
reached when solution in oil of turpentine alone 
becomes impracticable. 

At this stage, the practical mode of proceeding is 
to dissolve the fused resin in hot linseed-oil, afterwards 
to cautiously dilute this solution (whose temperature 
is maintained) with oil of turpentine. As the resin 
becomes more viscous (by degradation) precautions in 
diluting need to be increased ; at first cold oil of 
turpentine may be gradually introduced, later on it 
must be warmed, finally it must be raised to nearly 
the same temperature as the varnish. 

It will soon be observed that if the diluent is added 
a little too quickly precipitation of resin will ensue 
(this is a hint that dilution must be restricted) ; at 
the commencement of the diluting process, this pre- 
cipitate may, be again dissolved by constant mixing 
and the maintenance of temperature — no permanent 
harm will be done ; but such want of care must not 
occur towards the completion of dilution or the pre- 
cipitation will be permanent — ^the varnish will be 
spoilt ; the whole of the operations must be repeated 
with new materials. 

If the making of the varnish has been successfiilly 
completed, it must be mixed together as it cools. The 



106 Varnishes of the Italian Violin-makers. 

resulting cold varnish should be clear as crystal. 
Once completed and cooled, its viscosity can neither 
be increased nor diminished by evaporation or dilution. 
A few drops of cold turpentine added to it will imme- 
diately cause precipitation. Raised most cautiously 
to the temperature at which it was diluted, a little oil 
of turpentine of equal temperature is added ; while hot 
the operation appears to be successftd, but as the 
varnish cools it becomes turbid and breaks up (" le 
vemis louche"). 

The perplexing point is that as the solution of 
resin in oil becomes more viscous, appearing to 
require more oil of turpentine to bring it to a con- 
venient consistency, the constitution of the varnish 
demands a diminution of the diluent instead of an 
increase ; supposing always that the proportion of oil 
remains constant. 

Experts judge of the sufficiency of oil in a copal 
varnish by testing a drop of the oil solution cooled on 
a plate of cold glass ; this is an attempt to produce 
constancy of viscosity at this stage. The old violin- 
makers appear to have maintained the proportion of 
oil nearly constant, preferring to attempt to mani- 
pulate a thick varnish rather than diminish its e£Eect 
and brilliancy. 

The explanation suggested to account for these 
complex phenomena is that, in accordance with the 
usual laws, the solubility of the solid in the liquids 
increases with rise of temperature ; it decreases as 
the resemblance of its chemical composition to that 
of the solvents becomes less. As dehydration of the 
resin increases (of which viscosity and deepening 
colour are the signs), a point is soon reached when its 



Interesting Chemical Problems. 107 

solubility in oil of turpentine becomes very small ; 
linseed-oil then is the efficient solvent, oil of turpen- 
tine its diluent, both menstrua being miscible. When 
the quantity of solvent employed is greater than that 
required to produce a saturated solution, its dilution 
will be practicable to an extent dependent upon the 
margin of superfluity ; it will follow that, if with a 
gfiven resin convenient proportions of oil and essence 
are known, an increase of its dehydration, which 
reduces its solubility, will demand either an increase 
of oil or a decrease of essence : if the solvent remains 
a constant, the dilution must decrease ; if the quantity 
of diluent is constant, the solvent must increase. 

The complexity of the problem is aided by the fact 
that the small amount of solvent action of the oil of 
turpentine still further decreases as well as that of 
the more efficient solvent — linseed-oil. 

Since solution takes place at a comparatively high 
temperature, about ioo° C. and upwards, it may be 
supposed that, in many cases, oil-varnishes may be, 
when cool, supersaturated solutions ; it is known that 
they generally require mixing during the cooling 
process. 

This explanation must be admitted to be plausible 
even if its absolute correctness may be questioned ; in 
any case it cannot account for the fact that, after a 
varnish is successfully completed and cooled, its 
viscosity cannot be modified either by evaporation (if 
too thin) or by further dilution (if too thick) ; this 
critical condition must be quite apart from the ques- 
tion of saturation. If a varnish is too thick this 
defect may be remedied by making another varnish, 
under identical conditions, but with an increase of 



108 Varnishes of the Italian Violin-makers. 

diluent (a demonstration that in that particular case 
the point of saturation was not reached) ; when cool, 
the two varnishes may be mixed together. But if an 
attempt be made to add a little cold oil of turpentine 
to the varnish, precipitation ensues, or if it is most 
carefiilly raised to a proper temperature for dilution, 
and the wanting increment of diluent be most cau- 
tiously added, when the varnish cools it breaks up 
(" le vemis louche "). This is a mystery as to which 
experts oflFer no enlightenment. 

The observed phenomena make it quite clear that 
this critical condition of a varnish after completion 
cannot be dependent on solution ; it must result from 
some change in one or more of its constituents conse- 
quent on re-heating, or which has already taken place, 
and becomes apparent when this is attempted. The 
influence of moderate heat on the resin has already, it 
is hoped, been made suflBciently plain ; in many cases 
the solution of resin in oil has, previous to dilution, 
been heated for many hours at ioo° C. without affect- 
ing its constitution ; it is therefore most improbable 
that it should be materially affected by sufl5.cient 
re-heating (over a water-bath) to bring it again to 
about the same temperature. The part of the varnish 
which undergoes change or exhibits its effects on 
re-heating is manifestly the oil of turpentine used for 
dilution. It has already been demonstrated that a 
part only of this oil of turpentine is eventually spon- 
taneously evaporated from a pellicle of varnish ; it has 
been constantly observed that when oil of turpentine 
is mixed with a varnish at ioo° its characteristic 
odour soon disappears. The only reasonable . infer- 
ence which it appears possible to draw from all 



Interesting Chemical Problems. 109 

these facts is that, when the pinene is brought into 
intimate contact with the other ingredients of the 
varnish at ioo°, it undergoes some change of structure ; 
when re-heated it splits ; as it does when it is oxidized, 
to a very moderate degree, with nitric acid, and (after 
the completion of the oxidation reaction) is subjected 
to a temperature of ioo° or upwards, the distillate being 
a hydrocarbon which differs from the original dextro- 
pinene in being either laevo-rotatory or optically 
inactive, while the residue is no longer volatile. 
« « « « « 

As to the question of the old varnishes, I have 
been forced to abandon the old dogma in favour of a 
new one, which rests upon evidence so satisfactory in 
every direction that I adopt it without hesitation ; but 
as to the chemistry of the terpenes, while many of the 
doctrines taught by much esteemed teachers are now 
apparently untenable, no new faith supported by 
suflB.cient evidence has been formulated. 

M. Berthelot obtained derivatives from oil of tur- 
pentine which were often of two kinds : e,g,y artificial 
camphor, solid and fluid. He observed dichroism in 
some of his products, but offered no explanation. He 
derived laevo- from dextro-oil of turpentine by sub- 
jecting it to temperatures above 250° C. from 2 to 60 
hours (39, 10 and 11). He attributed these phenomena 
to isomerism. 

While isomerism is readily admitted, it seems 
insufficient to account for all the observed facts ; my 
poor efforts to ascertain the presence of isomerides in 
dextro-pinene have been unavailing; it appears to 
be perfectly homogeneous when well and carefiiUy 
rectified. 



110 Varnishes of the Italian Vtolin-makers. 

The facts which have come under my notice point 
rather to asymmetry in the terpenes or to phenomena 
which resemble those attending the inversion of 
sugar. 

A few of the points which appear to be significant 
are here briefly stated : — 

When oil of turpentine (dextro-pinene is always 
to be understood) is oxygenated it is invariably found 
to contain ozone. 

Oil of turpentine readily forms hydrates, but the 
derived hydrocarbon (terebene), whether Isbvo or 
inactive, does not form them (experiments which I 
have made on this point simply confirm those of more 
competent observers). 

When oil of turpentine is oxidized, even by the 
most gentle means, it is evident that (at least) two 
products are formed, the one resinous, the other fluid 
and volatile ; the tevo-rotation of the latter approxi- 
mates very closely to one-half of the dextro-rotation 
of the original pinene ; its aflB.nity for oxygen is quite 
different fi-om that of the original fluid, as may be 
easily observed by the behaviour of the two substances 
with nitric acid. The appearance of dichroism at the 
moment of separation of the component parts of the 
pinene seems peculiarly significant. ' 

In no case has it been found possible to transform 
the whole of the oil of turpentine either into resin or 
into an isomeric volatile oil: it always yields both 
solid and volatile products. 

The optical changes which may be observed in the 
terpenes are not more remarkable than those which 
affect the olfactory nerves. The odour of oil of tur- 
pentine is too characteristic to be mistaken; when 



Interesting Chemical Problems. Ill 

pinene is transformed into nitro-pinene this odour 
disappears and is replaced by another which is 
characteristic of pine- or fir-wood (here is a fact of 
singular significance) ; when the nitro-pinene is de- 
composed, the oxy-pinene has a pleasant aromatic 
scent ; on distillation, a fluid of an agreeable ethereal 
odour is obtained, the residue is aromatic. 

The mere mixing of oil of turpentine with a solution 
of oxy-terpene in linseed-oil, at ioo°, appears to deter- 
mine a change in its constitution ; its characteristic 
odour is changed, a part of it evaporates spoiitaneously, 
the residue persists in the pellicle and apparently 
becomes slowly oxidized. 

The changes in solubility, which have been described 
as occurring when oil of turpentine is oxidized by 
natural means (Mailand),when it is oxidized by nitric 
alcohol or by nitric acid, when the resulting oxide 
splits, when copal splits under the influence of heat, 
when colophony splits or is divided into two parts, 
i&c, are not without interest. 

All these observed transformations point to change 
of molecular construction, the determination of which 
is beyond the scope of the present research and of its 
author. It is to be hoped that experienced scientific 
chemists, having sufiicient time at their disposal, will 
take up the study of the terpenes and enlighten the 
scientific world on their constitutions and their 
mysterious modifications, natural and artificial. 



112 Varnishes of the Italian Violin-makers. 




Chapter VII. 
CONCLUSION. 

N order to form some definite conception of the 
general characteristics of the varnishes of the 
old violin-makers, the only practicable method (for 
persons who have not unlimited time at their disposal) 
is to study the descriptions given by experts, whose 
vocation has given them opportunities for examining 
many examples of the work of the different masters, 
or by musicians (professional or amateur) whose 
enthusiasm has led them to take an intelligent interest 
in the instruments on which they have played or which 
have come under their notice. 

Some quotations have already been given firom 
acknowledged authorities ; if these are read, with or 
without reference to the voluminous and discursive 
literature on the subject, one idea will be found to 
be generally prevalent — viz., that the old Italian 
varnishes, covering musical instruments, were some- 
thing special, different fi^om those which are found in 
the world on other articles ; that they had a common 
basis peculiar to themselves. Mr. Hart writes : — 
" Every instrument belonging to the school of 
" Cremona has it, more or less, in all its marvellous 
" beauty" (35). . . . " These varnishes " (Brescian, 
Cremonese, Neapolitan and Venetian) " are quite 



Conclusion. 113 



u 



it 
u 
u 



separable in one particular, which is the depth of 
" their colouring; and yet three of them, the Brescian, 
" Cremonese, and Venetian, have to all appearance 
" a common basis." . . . " If we examine the 
" Brescian varnish, we find an almost complete, 
" resemblance between the material of Gaspard di 
" Salo and that of his coadjutors, the colouring only 
" being different. Upon turning to the Cremonese, 
" we find that Joseph Guamerius, Stradiuarius, Carlo 

Bergonzi and a few others used varnish having the 

same characteristics, but again different in shade ; 

possibly the method of laying it upon the instrument 
" was peculiar to each maker. Similar facts are 
** observable in the Venetian specimens. The varnish 
'^ of Naples, again, is of a totally different composition, 
^' and as it was chiefly in vogue after the Cremonese 
" was lost, we ' may conclude that it was probably 
" produced by the Neapolitan makers for their own 
" need." (36.) 

There seems therefore to have been a continuity in 
the basis of the varnish fi-om the time of Gasparo da 
Sal6, in the early part of the sixteenth century, down to 
that of Stradivari and his pupils, in the middle of the 
eighteenth — ^roughly, for about two centuries. The 
Neapolitan was a later product ; it is not quite clear 
whether it is recognized as a relation of the family 
or not. 

If, as Mr. Hart supposes, this basis was a gum, 
common at the time, used for other purposes besides 
that of covering musical instruments, it is not a little 
remarkable that it should have been found in localities 
so far apart, politically and geographically, as Brescia 
and Cremona were firom Venice (communications were 

V. I 



114 Varnishes of the Italian Vtolin^makers. 

neither easy nor rapid in those days), and yet should 
not be mentioned by one of the authors quoted by 
M. Mailand in his research. 

Then Mr. Hart gives in a few words a description 
of the characteristic colours : — ** The Brescian is 

* mostly of a rich brown colour and soft texture, but 

* not so clear as the Cremonese. The Cremonese is 

* of various shades, the early instruments of the school 

* being chiefly amber-coloured, afterwards deepening 

* into a light red of charming appearance ; later still 

* into a rich brown of the Brescian tjrpe, though more 

* transparent, and frequently broken up, while the 

* earlier kinds are velvet-like. The Venetian is also 
'of various shades, 'chiefly light red, and exceed- 
' ingly transparent. The Neapolitan varnish (a 

* generic term including that of Milan and a few 
' other places) is very clear, and chiefly yellow in 

* colour, but wanting 'the dainty softness of the 
' Cremonese." (38.) 

Reference to the writings of other authorities 

enables some of the details of this sketch to be filled 

in. The Messrs. Hill write : — " We may thus fairly 

" assume that Stradivari was at an early stage of his 

" career initiated into the traditions and methods 

" practised by his master and predecessors. Once 

" freed from his connection with Amati, we see him 

** seeking, by changes effected in the colours of his 

" varnish, to give a different appearance to his instni- 

" ments, as is evidenced by his departure from the 

" hitherto conventional Amati yellow. We say 

" * conventional,' because, throughout the four genera- 

" tions of that remarkable family, every member of 

" it kept, with rare exceptions, to the same tint of 



Conclusion. 115 



" colour." (Stradivari, 1 74.) We learn from this 
that the Amati used varnish of a yellow colour for 
four generations ; that Stradivari did not depart from 
the old traditions, as to varnish, until his connection 
with Nicol6 Amati was broken by his death in 1684. 
Stradivari designed the "LongStrad" in 1690; he 
returned to Amati traditions, as to form, in 1698. It 
is as rational to assume that Stradivari changed his 
varnish for the sake of the appearance of his instru- 
ments as it would be to assume that he changed their 
form for the same reason ; a motive must be sought 
which is more consistent with the character of this 
eminent man and with the steady purpose of his life 
and work. 

M. F6tis fixes for us the epoch of change, and also 
expresses in felicitous terms his view of the aim of his 
long life. He writes (the translation is mine) : — 
" In 1 700, the artist has reached his fifty-sixth yekr. 
" His talent is then in its fiill strength ; the instruments 
" which leave his hands, from this time up to 1725, 
** are so many perfect works. He no longer feels his 
" way. Certain of what he does, he carries into the 
" smallest details the most beautifiil finish. His 
" model has all desirable amplitude ; he draws the 
" contours of it with a taste, a purity, which for a 
" century and a half have excited the admiration of 
" connoisseurs." ..." The beautiful warm tones 
" of the varnish of Stradivari date from this epoch. 
" Its body is fine and of great elasticity." (Vol. VIII., 

152.) . . . 

" Stradivari belonged to that small number of 
" those eminent men (hommes d'61ite) who, setting 
" up for themselves as their end and aim, perfection 

i2 



116 Varnishes of the Italian Violin-makers. 

* (so far as is given to humanity to attain it), do not 
' allow themselves to stray from the path which may 

* lead them to it ; whom nothing distracts, nothing 

* deflects from their object ; whom deceptions do not 

* discourage, and who, full of faith in the value of 

* their object as in their abilities to realize it, inces- 

* santly recommence whatever they have done well, in 

* order to arrive at the best possible. For Stradivari, 

* violin-making was the whole world; he concentrated 
' in it the whole of his personality. This is the way 

to rise to eminence, when aptitude answers to 

will." (153.) 
The distinguishing attribute of Stradivari, after 
he had passed his youth and had reached the prime 
of his mental capacity, is his thoughtftilness : no 
detail of form or of construction escaped the most 
minute consideration. To suppose that he overlooked 
the influence of varnish on the tone of his instruments, 
as well as on their appearance, can only arise from 
ignorance of the chemical and physical problems 
involved ; it may be taken as certain that he devoted 
time and thought to the study of this question as well 
as of all others that were factors affecting the attain- 
ment of his ideal perfection. 

After 1725 the varnish of Stradivari tended more 
towards shades of brown (this change could not have 
been "viHithout reason) ; this tendency is observable not 
only in his instruments, but also in those of other 
makers living at that time, some of whom were his 
pupils and others most probably cognisant of his 
opinions. A quotation has already been given from 
M. Vidal in which he describes the varnishes of the 
pupils of Stradivari as more heavy, thick, and red- 



Conclusion. 117 



brown. In his opinion, this is an indication of the 
commencement of decline. 

It will now not be difficult, from the descriptions 
and opinions of the various experts, to construct 
the probable course of evolution of the Brescian- 
Cremonese varnish ; the makers of these two towns 
(not far distant from each other) appear to have had 
much mutual influence. The earliest examples of 
Brescia were brown, a little wanting in transparency ; 
the earliest Cremonese were mostly yellow or amber- 
coloured, more transparent than those of Brescia. 
Deeper colours began to appear some time after 1684 ; 
the warm tones of Stradivari about 1700; after 1725 
brown shades (which do not seem at any period tp 
have been entirely absent) were again introduced ; 
the varnish finally disappeared about 1750. The 
Brescian varnishes were very soft ; those of Cremona 
had a " dainty softness ; " some of the varnishes of 
Stradivari appear to have had a tendency towards 
hardness — were a little " chippy." 

The Venetian varnishes were of various tints, 
resembling those of Cremona— often light red. No 
facts as to their evolution are available; but, ap- 
parently, they excelled all others in transparency. 
Mr. Hart's expression is " exceedingly transparent." 

The Neapolitan varnishes were mostly yellow, but 
(as I know from my own experience) they were some- 
times of a red-brown colour, which is more garish than 
the warm tints of the Cremonese and Venetian types. 
The yellow varnishes were " very clear," but the red 
ones were wanting in perfect transparency. In 
general they were harder than the Cremonese and 
Venetian. 



118 Varnishes of the Italian Violin-makers. 

The assumption that the common basis of all the 
varnishes was a gum is rejected ; it is assumed that 
the old violin-makers used as the resinous ingredients 
of their varnishes the products of the coniferous trees 
which g^ew in their vicinity. 

A simple and almost obvious explanation is sug- 
gested for the slight difference between the Brescian- 
Cremonese and the Venetian varnishes : — It is well 
known that turpentine derived from larch (Larix 
Europaea) has for centuries been an article of com- 
merce from Venice; larch products are therefore 
allotted to Venice, and those of the pines and firs 
(Pinus, Abies, and Picea) to Brescia and Cremona. 
We adopt the purely arbitrary distinction of calling 
varnishes derived from fir- or pine-turpentine, gali- 
pot, rosin or colophony — Cremonese ; those derived 
from Venice turpentine — ^Venetian. We soon find, 
by experiment, that the characteristics of these two 
kinds of varnish correspond closely with those ascribed 
to the two districts. Both yield similar varieties of 
colour, but there is a small difference in shade — the 
Venetian is slightly more transparent, and is a little 
hctrder when perfectly dry. 

M. Mailand cites only two authors who wrote in 
the sixteenth century, Alexis the Piedmontese (1550) 
and Fioravanti (Bologna, 1564). Piedmont is a little 
remote from our scene ; Bologna is nearer. Alexis 
prescribes " turpentine of Venice " ; also " pine resin, 
fat and white" (galipot). Fioravanti prescribes pine 
oil (perhaps crude turpentine, not oil of turpentine as 
M. Mailand suggests) ; Grreek pitch (resin from the 
pine trees of Calabria) ; pine resin. At the present 
time there is no difiiculty in obtaining turpentine of 



Conclusion. 119 



Venice, which is still in commerce; instead of the 
vaxious pine products of Italy, it is now more convenient 
to use those in commerce which come from ports in or 
near the Gulf of Mexico (from regions where the 
climate much resembles that of Italy) — viz., colophony 
and oil of turpentine. 

In order to obtain resins suitable for the production 
of fine varnish, these various substances require oxida- 
tion. It requires no advanced student of chemistry 
to discern that the varnishes on the old instruments 
are products of oxidation ; thus much I had settled in 
my mind many years before this research was con- 
templated. What oxidant was employed in the old 
days there is no means of knowing. That which has 
been found to be eminently suited to the purpose is 
nitric acid, which has the inestimable advantage of 
leaving nothing behind it — it parts with a portion 
or the whole of its oxygen to the turpentine or colo- 
phony; the residuary nitrogen, or its lower oxides, 
and water are easily perfectly eliminated by heat. 

Nitric acid, as the oxidant, affords a simple and 
complete explanation for the evolution of all the 
different t5^es of varnish. Some other oxidant may 
enable an equally plausible explanation to be given, 
but I must confess that I am not able to suggest 
what it could be. 

There can be no possible doubt that nitric acid 
(called then spirit of nitre) was easily obtainable in 
the sixteenth century ; it had been known since the 
eighth. It was in those days derived from nitre 
(potassic nitrate), nitrate of soda being not then 
abundantly available as it is to-day. 

The oxidation of the turpentines is indispensable ; 



. e — UL. 



120 Varnishes of the Italian Vtolin-makers. 



the source of the oxygen is in no way material. 
Turpentine may have been found in Italy sufficiently 
oxidized by natural processes ; the Greek pitch from 
the conifers of Calabria, mentioned by Fioravanti and 
Bonanni, is possibly a product of this kind. The 
oxygenating powers of oil of turpentine may have 
been utilized ; if so, it must have been a slow process 
indeed. Finally, nitric acid or some other artificial 
oxidant may have been employed. The discovery of 
the oxidant actually used would be most interesting ; 
in the meantime, nitric acid is so satisfactory in every 
respect that it cannot easily be excelled. 

When Gasparo came from Salo to Brescia he must 
have known resin or colophony, for it is essential to 
the friction of the hair of a bow on a string. Very 
probably this substance was the first used for a varnish 
also. 

If a good sample of colophony be oxidized to a 
moderate extent, a brilliant resin of an intense yellow 
colour, very much resembling gamboge, will be 
obtained. This resin is completely soluble in oil of 
turpentine ; consequently a varnish may be produced 
from it without any admixture of linseed-oil. When 
an instrument has been covered with this varnish, its 
appearance will be satisfactory in every respect: 
the ground tone will be amber-yellow, the refractions 
caused by the grain of the wood will be of a red 
colour; but when this varnish dries (in a month or 
six weeks), it will be found to be wanting in sufficient 
cohesion to bear wear or polishing. 

There are two obvious remedies for this defect ! 
the first is to introduce into the varnish a proportion 
of thick raw turpentine ; the second, a similar intro- 



Conclusion. 121 



ductipn of boiled linseed-oil. With the experience of 
the vamish-makers of a century at my disposal, I did 
not hesitate to adopt the second expedient as the best; 
but it by no means follows that, had I lived in the 
sixteenth century, I should have taken the same 
course; from the facts at our disposal, it would appear 
that the early makers of Brescia adopted the first 
expedient. 

The most obvious means, then, for forming a 
varnish from the yellow resin would be to dissolve it 
in crude turpentine by the aid of a charcoal brazier or 
a similar direct-heating appliance, keeping the in- 
gredients well stirred with a wooden spatula or rod. 
The effect of this operation would be to cause dehydra- 
tion, as well as a little decomposition, of the yellow 
resin; the resulting varnish would be of a brown 
colour — ^pale, if the heating had been moderate ; of a 
deeper shade if the heating had been extended in 
duration or in degree. If the varnish were made cold, 
which would require some time, the colour would be 
yellow. 

When an instrument had been covered with this 
transparent yellow or brown varnish, and after the 
volatile part of the turpentine had slowly and spon- 
taneously evaporated, the remaining pellicle would be 
a pure resin, modified by oxide of turpentine ; such a 
varnish would exactly answer the description given by 
M. F6tis of the varnish of G. P. Maggini — " a pale 
brown spirit varnish " : for it would be completely 
soluble in alcohol. 

It will be easily understood, from what has been 
already stated, that such a turpentine varnish would 
be more transparent if pale. 



122 VamtsAes of the Italian Violin-makers. 

These turpentine varnishes represent the early 
varnishes of Brescia (perhaps of Cremona). 

It is conceivable that in the hands of men of the 
high intelligence of the old Brescian and Cremonese 
masters, whose powers of observation had been culti- 
vated and increased by long study of technical (not to 
say scientific) problems, the defects of the method of 
production and of the finished varnish would soon 
become apparent. The introduction of linseed-oil 
would be the first obvious improvement in the dura- 
bility and cohesion of the pellicle. It has already 
been demonstrated by M. Mailand that methods for 
the necessary transformation of raw into boiled oil 
were known in the time of Maggini. 

This improvement, still accompanied by the ap- 
plication of direct heat, would explain the manu- 
facture of oil-varnish such as that used on the later 
instruments of Maggini ; on those of the earlier makers 
of Cremona ; and, up to a much later date, on those of 
Italian makers, who were not very particular about 
the colour and quality of their varnish. The tints 
would be varying shades of golden brown with red 
and red-brown refiractions. 

The uncertainty of result caused by difficulty in 
controlling the direct mode of heating must soon 
have led to the use of the water-bath (the " balneum 
marise" of the old chemists). It is impossible to say 
how early this simple device was known ; Alexis pre- 
scribes its use in 1550; Zahn in 1685. The introduc- 
tion of this improvement would lead at once to greater 
certainty in the manufacture : degree of temperature 
(100°) would now be a constant, the factor movable 
at will, duration. Not only this, the brown colour, 



Conclusion. 123 



due to the action on the varnish of the over-heated 
surface of the containing vessel, would disappear ; 
instead of the brown tones of the early period, the 
yellow and orange colours, with red refractions, of 
the Amati, would now appear ; the ultimate produc- 
tion of the warm tones of Stradivari's great epoch 
would be a question of evolution. From the exami- 
nation of about a dozen instruments of Stradivari and 
from the representations of a few others, I conclude 
that this painstaking workman invariably used the 
water-bath in preparing his varnish. 

The influence of the colour and character of the 
wood on the apparent colour of the varnish is greater 
than one could conceive possible. It is intelligible 
when the effect of complicated reflection, from vary- 
ing planes, is well understood; but even then the 
practical results are often surprising. 

The effect of these varnishes on the wood of the 
belly or table is eminently characteristic. In forming 
the moulding of this important part of the instrument 
(especially of the smaller ones), the angle of section of 
the grain is constantly changing, thus bringing into 
view, more or less, the medullary rays. Moreover, 
the absorption of varnish is greater where the tiny 
cylindrical fibres have been cut across. This varia- 
tion of section, under the refracting varnish, produces 
the mottled appearance which Mr. Hart so truthftiUy 
describes. The beautiftil satin-like appearance of 
some of the tables is remarkable, considering that 
they are made of wood so apparently devoid of 
character as that of Abies or Picea. 

From what has been said, it will be understood 
how it happens that, among the works of the old 



124 Varnishes of the Italian Violin-makers. 

masters^ no two instruments are alike. The variations 
in the varnish itself are infinite (it seems unlikely 
that they prepared any considerable quantity at one 
time) ; difference in the proportion of nitric acid, in 
the duration of nitrification, in the duration and 
method of heating, must have produced unlimited 
variety. The effect was further greatly modified by 
the quality and character of the wood. 

No doubt age has, to a small extent, changed the 
colour of the old instruments; but this change, which 
is not material, is not to be attributed to modification 
of the varnish, but to the wood. All woody tissue has 
a tendency to become brown with ag e e ven the prac- 
tically pure cellulose of paper has a decided tendency 
in this direction. The apparent colour of the old 
instruments is therefore to-day somewhat deeper than 
when they left the hands of their makers; always 
supposing that the thickness of the pellicle has re- 
mained practically constant, and that it has not been 
unduly exposed to direct sunlight. 

If this hypothetical and manifestly imperfect sketch 
(for how little reliable material is to be found when it 
is tested !) of the old varnishes (fi-om their origin until 
Stradivari had reached his prime) has any solid 
foundation, it may be supposed that when once in 
the course of this research the true starting-point had 
been reached, the evolution of the varnishes, in my 
hands, would follow the same course as in the old days. 
The Italians were working in the dark — ^they could 
not have known the significance of their simple opera- 
tions ; while I had some knowledge of modem chemi- 
cal laws and theory to enlighten my task. For this 
reason — not from any inferior natural acumen — ^the 



Conclusion. 125 



slow evolution in their hands extended over two cen- 
turies, while in mine it occupied only as many years. 
As I proceeded, every varnish which indicated a new 
phase of evolution was laid on an instrument, to 
enable progress to be clearly appreciated. The modern 
evolution is thus illustrated by a collection of little 
more than a score of instruments ; the ancient one was 
spread over many thousands. 

In order that my readers may be able to follow 
the gradual evolution and appreciate its significance, 
firom the earliest Brescian varnishes down to the 
masterpieces of Stradivari, I will give, as concisely 
as possible, the exact dates on which the different 
varnishes were prepared. The different processes have 
already been sufficiently described in detail. 

The Neapolitan varnishes will be intentionally 
omitted, because their mode of production became 
prematurely known to me from circumstances which 
have already been fiiUy explained. 

Although dichroism had been observed in May, and 
the identity of the colours of oxidized pitch-pine with 
those of the old varnishes had been discerned in June, 
it was not until the end of November, 1900, that 
experiment and investigation enabled me to conclude 
with certainty that these phenomena were due to the 
presence of oxides of turpentine and their derivatives. 
This preliminary investigation was not without in- 
terest, but the facts brought to light are scarcely 
relevant to the present question. 

The preparation of turpentine varnishes (without 
linseed-oil) extended from the 30th December, 1900, 
to the 19th January, 1901. The earliest of these var- 
nishes were yellow ; the later yellow-brown. These 



126 Varnishes of the Italian Violin-makers. 

represent the earlier varnishes of Brescia, which 
M. F6tis supposed to be spirit varnishes ; the pellicles 
were perfectly soluble in aqueous alcohol (about 90 
per cent.). 

After testing the last of these varnishes, I con- 
cluded that they were wanting in cohesion; I therefore 
determined on the admixture of siccative linseed-oil. 
It is now a matter of great regret to me that I removed 
these pure turpentine varnishes from the violins which 
I covered with them ; slips of wood varnished with 
them, at the same time, show that the varnish has 
gfradually solidified. 

The first oil-varnish was made from colophony (in 
the dry way), over a sand-bath, on the 31st January, 
1901 ; a. perfectly successfiil oil-varnish made in this 
way was prepared on the 13 th February, 1901. A 
violin was coated with this varnish ; the colour is a 
peculiar golden yellow-brown, with brown shadows 
and red-brown refractions ; it very much resembles 
that of the " Dumas " violin made by G. P. Maggini, 
the back of which is represented in colours in the 
book of Messrs. Hill (Maggini 61), but it does not 
exactly correspond with the tints of the drawing. 

Several other varnishes were made in the same 
way with the sand-bath ; but I soon realized that 
this mode of heating was too uncertain. Some fine 
varnishes were made, others were spoilt by too high a 
temperature ; in all cases they were of a brown tone. 
Deep-brown varnish made with the sand-bath requires 
too much oil for solution to suit my ideal. This mode 
of heating was abandoned ; henceforth the water-bath 
and water-oven were used instead. 

The first oil-varnish (made in the wet way), from 



Conclusion. 127 



colophony dissolved in oil of turpentine, with the 
water-bath, was prepared on the 19th February, 1901. 
A violin was coated with this ; it is pale yellow with 
light red refractions. The quality leaves nothing to 
be desired, but it is rather too pale to be effective. 

On the 2 1 St February, 1901, a similar varnish was 
made from turpentine of Venice ; a violin was covered 
with it. A varnish of the finest quality — golden 
yellow, with red-brown refractions. This Venetian 
varnish is much more effective than the corresponding 
Cremonese; as a pale varnish, it leaves absolutely 
nothing to be desired. 

On the 17th March, 1901, a varnish was made from 
colophony dissolved in oil of turpentine, similar to the 
last Cremonese varnish ; it was heated over a water- 
bath for five hours. The increased depth of colour 
and viscosity were observed ; it was conjectured that 
the changes caused by this heating signified " dehy- 
dration." This varnish was used for a violin and a 
viola; it is a good varnish of the Amati type — 
perfectly transparent, brilliant, golden yellow with 
red-brown refractions; agreeing exactly with the 
description which Cartier gave of that on the violins 
made by Andrea Amati for the King of France, which 
has already been quoted; its tint exactly matches that 
of early varnishes of Stradivari. 

Considerable time was now spent on endeavouring 
to determine interesting theoretical questions. 

On the 27th April, 1901, a Cremonese varnish was 
made similar to the last, but of a deeper colour (oxida- 
tion and dehydration both increased). A violin and 
violoncello were covered with it ; coloiu", orange with 
red-brown refractions. 



128 Varnishes of the Italian Vtolin-makers. 



On the 24th May, 1901, a Venetian varnish Avas 
prepared, which was used for two violins, the one of 
which exhibited a fine orange-red colour, the other 
inclining more to orange-brown shades. The reason 
for this difference was that the first violin had been 
previously covered with a Cremonese varnish, which 
had been removed by alcohol ; the second was a virgin 
violin. A very transparent varnish of perfect quality, 
it solidifies to a slightly harder pellicle than a corres- 
ponding Cremonese varnish. 

On the 4th September, 1901, a Cremonese varnish 
(C), a Venetian varnish (V), were made on precisely 
similar lines ; they were both used for violins, both 
exhibiting the orange-red tones of Stradivari. Both 
varnishes were of excellent quality, but the Venetian 
was judged to be slightly superior in colour, trans- 
parency, and brilliancy. The Venetian pellicle was 
decidedly harder than the other. 

On the 29th November, 1901, a Venetian varnish 
was made, and a violoncello was covered with it. 
This varnish answers the description given by M. de 
Try of that on the Stradivari violoncello of 1725, on 
which he had played in Madrid ; it is a very trans- 
parent varnish, of an amber-yellow colour with bright 
red refiractions. 

At this stage I observed the effect caused by pro- 
longed nitrification on oxidation. 

On the 9th December, 1901, two Cremonese var- 
nishes were made precisely alike up to the oxidation 
stage (nitrification extended over four days) ; the one 
was subsequently heated for a short time only in the 
water-oven, the other for six hours ; the one was not 
perceptibly, the other was considerably, dehydrated. 



Conclusion. 129 



One was used for a violin, the other for a viola. That 
on the violin is of a bright yellow colour with copper- 
coloured refractions ; the dehydrated one on the viola 
is deep orange with red-brown refractions. Both are 
v^ry good, but perhaps not quite so effective as 
Venetian varnishes. 

Finally, on the 25th April, 1902, a varnish was 
made from equal parts of Venice turpentine and oil of 
turpentine. Nitrification was extended over three 
weeks ; consequently oxidation was probably the 
maximum possible at one operation. A violin was 
covered with this varnish : a transparent, brilliant 
varnish of a deep orange-red, the most coloured 
Venetian varnish in my collection (it was, of course, 
hybrid, for a considerable part of the oil of turpentine 
was oxidized). 

As far as I am aware, this last varnish exhausted 
my powers of evolution (not of variety) of Cremonese 
and Venetian varnishes, consistent with the mainten- 
ance of my ideal of quality ; of varnishes of declining 
quality I shall write later. 

In the month of August, 1902, the Messrs. Hill 
published their book on Stradivari ; my copy reached 
me about the twenty-fifth day of that month. This 
book contains excellent representations of some of the 
finest instruments of Stradivari in existence. After 
examining these coloured drawings (since I have 
practised water-colour drawing for nearly half a cen- 
tury, my eye-^whatever natural defect it may have — 
cannot be considered untrained), I sought in my col- 
lection of instruments for examples which might 
possibly match them. Somewhat to my surprise, I 
found that, in almost every case, I possessed instru- 

V. K 



130 Varnishes of the Italian Violin-makers. 

ments which corresponded exactly, or very closely 
indeed, with the tints of the drawings. It is necessary 
that I should express myself clearly on this point : 
what I mean to assert is that, if I gave a commission 
to a water-colour artist to depict my instruments, I 
should consider that he acquitted himself very well 
indeed if he came as near as the tints in the book. 
The backs only of the instruments were taken as a 
guide ; the tables are always more or less modified by 
resin, &c. 

Here are the instruments of Stradivari and the 
dates they bear, with the corresponding dates on 
which I made the varnishes, the tints of which match 
the plates : — 

Long Pattern Strad. Violin 1693 Page 48 17 March, 1901. 

Inlaid Viola 1696 „ 108 17 March, 1901. 

Tuscan Strad. Violin 1690 „ 164 (C) 4 Sept., 1 901. 

Duport Strad. Violoncello 171 1 „ 144 24 May, 1901. 

Rode Strad. Violin 1722 „ 238 (V) 4 Sept., 1901. 

Alard Strad. Violin 1715 ,, 90 25 April, 1902. 

It will be also remembered that M. de Try described 
a varnish on a Strad. violoncello of 1725, the similar 
varnish to which I made on the 29th November, 1901. 
Soon after, I acquired a copy of Mr. A. J. Hipkins' 
book, and on applying the same test to the drawings 
therein I found an exact match for the 

HeUier Strad. Violin .... 1679 Plate XXV. 27 April, 1901. 

The Alard Strad. violin and the King Joseph 
Guamerius del Gesii violin, on Plate XXVI., I am 
unable to match ; although, curiously enough, I can 
match the Alard Strad. violin as depicted by the 
Messrs. Hill. 

Of the instruments pictured by the Messrs. Hill, I 
had seen one, the " Tuscan Strad." violin, about fifteen 



Conclusion. 131 



years before, I possessed their coloured illustration 
(dated 1 889) of this instrument ; the others I had never 
seen, neither was I aware of their existence. 

It has already been distinctly stated that it is 
impossible for anyone to know beforehand the exact 
colour which an instrument will exhibit when coated 
with a varnish (such as those whose method of pro- 
duction I have described) ; consequently, there can 
be no question here of copying what I had never 
seen ; the matching of the tints is a pure coincidence. 
In the case of my Venetian varnish of the 24th May, 
1 90 1, as I have related, two violins were covered with 
this varnish, one of which matched the colour of the 
Duport violoncello, the other not resembling it. 
The reason for the great difference in colour between 
these two violins, covered with the same varnish, did 
not for some time occur to me. I attributed it at first 
to the different character of the wood ; afterwards I 
remembered that one of these violins had been coated 
with a Cremonese varnish, which, not being entirely 
satisfactory, was removed by the aid of methylated 
alcohol. This was the violin which, when afterwards 
covered with the Venetian varnish, matched the colour 
of the Duport violoncello. Now, although it is not 
very difl&cult to remove a varnish firom the surface of 
an instrument, that with which the wood is saturated 
cannot be removed; consequently, when this wood 
was subsequently coated with another varnish, the 
result may be said to constitute a kind of hybrid 
between the Cremonese and Venetian varieties, the 
latter largely preponderating. The soundness of this 
hypothesis was demonstrated, in March, 1903, when I 
prepared a varnish whose turpentine constituents 

k2 



132 Varnishes of the Italian Violin-makers. 

were loo parts turpentine of Venice, 17 parts of oil of 
turpentine, 5 parts of colophony; with this hybrid 
varnish I coated a violoncello whose wood resembled 
that of the Duport Strad., and, as I expected, I ob- 
tained an. almost exact replica of this instrument. 

My varnish of the 25th April, 1902, was made from 
equal parts of turpentine of Venice and of oil of 
turpentine (a sample of a yellow tinge that would be 
sure to yield a considerable proportion of resin by 
oxidation) ; this was therefore again a hybrid varnish, 
which matched the colour of the Alard Strad. violin 
of 1715 (Hill). 

Attention is not called to these facts from a desire 
to prejudge the question of the identity of the modern 
with the ancient varnish, but because the different 
tints revealed interesting deductions which were 
totally unexpected, which will now be explained. 

If it be assumed, for the present purpose, that 
Stradivari was experimenting with varnishes identical 
with the modem ones, it is not at all difficult to 
classify the varnishes on the various instruments 
(judging only by the tints) : — 



The Long Pattern Violin (1693) 
The Inlaid Viola (1696) 
The Tuscan Violin (1690) 
The HiUier Violin (1679) 

The Rode Violin (1722) 

The Madrid Violoncello (1725) 



Cremonese vamish. 



) 



Venetian vamish. 



The Alard Violin (1715) 

The Duport Violoncello (171 1) Cremonese and Vene- 

The Betts Violin (1704) 



tian hybrid vamish. 



Conclusion. 133 



While engaged in these experiments, I have ex- 
amined about a dozen of Stradivari's instruments 
(violins and violas), the varnishes on which were 
Cremonese and hybrid. I did not observe one pure 
Venetian (one or two of the varnishes had been 
tampered with) ; they were certainly all dichroic, not 
one of them was coloured. 

Not only is there a difference in tint between a 
Cremonese and Venetian varnish, but there is a most 
decided difference in the tone which is obtainable by 
the impregnation of the wood by the two varieties. 
An instrument treated with a Venetian varnish will 
possess a tone more resonant, powerful, and brilliant . 
than that obtainablie from a precisely similar instru- 
ment covered with Cremonese ; the derivatives of 
turpentine of Venice appear to give more resistance 
to the wood which is saturated with them ; the wood 
of pitch-pine is much more elastic than that of larch. 

In contemplating these facts and deductions, it is 
impossible to avoid reference to the life and work of 
Stradivari. In the light which has so far been thrown 
on his career, his aims are sometimes not intelligible ; 
but this novel view gives a possible explanation of 
facts which other writers have observed and recorded. 

Down to the time of Stradivari, if we may judge 
by the occasional recurrence of brown varnishes, it 
may be assumed that the Cremonese makers employed 
some crude or direct method of heating in making 
their varnishes ; if not generally, at least frequently. 

All the varnishes of Stradivari which I have seen, 
or of which I possess illustrations, were, without doubt, 
made with the water-bath. The first improvement of 
varnish which Stradivari adopted was therefore the 



134 Varnishes of the Italian Violin-makers. 

use of the water-bath ; if he did not introduce it, he 
recognized its great advantage, and appears to have 
invariably availed himself of it. 

Up to about the year 1696 we find Stradivari 
experimenting with the form of his instruments. The 
*• Long Strad." was introduced in 1690 ; he returned to 
Amati traditions in 1698 (Hill). What was he 
seeking ? The softness, equality, purity and volume 
of tone of the instruments of his predecessors and of 
his own early life left nothing to be desired. It seems 
evident that what he sought was greater power and 
brilliancy. Was not this what he expected to obtain 
from the " Long Pattern " ? Were not experiments 
with curves of smaller radii for table and back (such 
as are to be observed in Stainer's instruments) in the 
same direction? Why did he return to Amati tra- 
ditions (as to form) in 1698 ? 

If we may judge from the small amoimt of evidence 
before us, he was still using Cremonese varnish in 
1696 {vide Inlaid Viola). He had exerted his powers 
in hairdening (as far as possible) the Cremonese 
varnish, and had produced the Tuscan Violin (1690) ; 
he could not possibly have gone beyond this varnish 
in oxidation and dehydration without necessitating 
an increase of oil, which would obviously defeat his 
object. 

Not long after 1696, the idea of trying the effect 
of Venetian varnish (/>., turpentine of Venice) appears 
to have occurred or to have been suggested to him 
(the influence of Montagnana appears possible, but 
there are no means of obtaining any light on this 
point ; so little is recorded of the " great Venetian " 
that whether he was of Venetian origin or was a 



Conclusion. 135 



Cremonese who migrated to Venice, there is no 
evidence to show). Then he not only obtained 
different tints of colour ; but the use of Venetian var- 
nish, by imparting increased resistance to his dia- 
phragms, gave him the qualities of brilliancy and 
power which he had so long soi;ight in vain. 

His experiments with form and thickness of wood 
seem from about 1698 to have ceased; during his 
great period (1700 — 1725) he appears to have devoted 
his attention to the effects of different combinations of 
Cremonese and Venetian varnishes— of hybridisation. 
Even so late as 1725 we find him using a pure 
Venetian varnish — if the surmise as to the Madrid 
violoncello described by M. de Try is well' founded. 

It must be evident that this change of varnish may 
account for the preference that is given, for large 
rooms, to the instruments of the pupils of Stradivari 
and such of his contemporaries as may have acquired 
the necessary information, — ^to those of the makers that 
constitute the Stradivari group or school. The im- 
provement in tone cannot be traced to any consider- 
able change in form or wood. The form of the 
instruments of Joseph Guarnerius del Jesfi, for 
instance, differs considerably from that of the 
Stradivari instruments ; but we find a similar brilliancy 
of tone. Copies of the Guarnerius instruments covered 
with Maggini or Amati varnish lack this brilliancy 
of tone, as far as my experiments with such violins go. 

The remarks of Mr. Hart are mbst significant in 
this connection. Observe what he says about the 
colours : — " The Cremonese is of various shades, the 
" early instruments of the school being chiefly amber- 
" coloured, afterwards deepening into a light red of 



136 Varnishes of the Italian Violin-makers. 

" charming appearance, later still into a rich brown 
" of the Brescian type, though more transparent/' 
....** The Venetian is also of various shades, 
" chiefly light red, and exceedingly transparent/' 
He had previously made this remark : — " Upon 
" turning to the Cremonese, we find that Joseph 
" Guamerius, Stradiuarius, Carlo Bergonzi, and a few 
** others, used varnish having the same characteristics, 
" but again different in shade/' It is remarkable 
that he uses precisely the same expression, "light 
red," to describe the Cremonese at a particular period 
as he does for the Venetian ; then he puts the " light 
red " into its proper position between amber-coloured 
and the recurrence of brown ; finally he distinctly 
discriminates the varnish of the Stradivari group. It 
has already been suggested that Stradivari used the 
old amber-coloured varnish as lately as 1696 (Inlaid 
Viola), and that soon after that date he began to use 
Venetian; it will be presently shown that after 1725 
his varnish became more brown, and that his pupils 
began to push dehydration, the consequent production 
of brown colour, too far. 

To judge by my brief experience, instruments 
treated with pure Venetian varnish are most brilliant, 
resonant and powerftil, but the tone is not very easy 
of perfect production ; for amateurs, perhaps, not very 
agreeable to produce. On the other hand, instruments 
covered with the Cremonese t3rpe give out their tone 
with a facility and a fireedom which are delightful to 
most amateurs. The artiste prefers to feel that resist- 
ance to his bow which the Venetian varnish imparts ; 
with a little experience, even the advanced amateur 
soon lays aside the more facile instrument. Stradivari 



Conclusion. 137 



appears to have sought a via media ; he must have 
numbered more amateurs than professionals among 
his clients. 

The evidence given by the small number of instru- 
ments in question is capable of the interpretation that 
Stradivari began the use of turpentine of Venice ten- 
tatively, at first mixing it with pine-turpentine so as 
to produce hybrid varnish, and that he did not try the 
effect of pure Venetian varnish until after 1720; this 
would be consistent with the general trend of evolu- 
tion of the varnishes of Cremona towards increased 
hardness and solidity. 

Some writers have attributed the increased reson- 
ance of the instruments of Stradivari's great period to 
the quality of the wood of construction. Without a 
long experience, it is impossible to give a decisive 
opinion ; but I am strongly disposed to attribute the 
superior qualities of his instruments (1700 — 1725) 
mainly to the varnish which he then used. 

I varnished two violoncellos (by different makers), 
one with a pure Cremonese, the other with a pure 
Venetian varnish ; the models of the two instruments 
are practically identical, the wood of the latter is 
perhaps a little thicker than that of the former. I 
have played on one or other of them daily for more 
than a year; my opinion remains precisely what it 
was on the first trial. The difference in tone can be 
easily distinguished ; the player feels the difference 
in resistance to the bow. The resonances of the 
Venetian instrument are incomparably superior to 
those of the Cremonese ; it is also far more brilliant 
and powerful ; the bass strings are particularly fine. 
It seems to me inconceivable that the striking dif- 



138 Varnishes of the Italian Violin-makers. 

ference in tone can with justice be attributed to the 
wood ; certainly the superiority of the one instrument 
does not arise from any want of care or skill in the 
cabinet-work of the other. From violinists, moreover, 
I have found that, in their opinion, violins, identical 
in every respect, have varpng characteristics of 
tone imparted to them by Cremonese and Venetian 
varnishes. 

« « « « « 

Having traced the progress of evolution towards 
its zenith of perfection, as exhibited in the finest of 
the warm-toned (colour) instruments of Stradivari, 
it will now be necessary to trace the evolution of 
decline. 

When the old Italian varnishes came into existence, 
the exigencies of life under the genial climate of 
Italy were such that it was of little moment that 
an instrument -maker was obliged to wait for 
many months before he could realize, in a current 
form, the results of his mental and physical labour. 
It was not until after Antonio Stradivari had passed 
his prime that change in this respect began to be 
felt. This admirable workman could afford (as some 
of his biographers suppose) to spend twenty years 
of the best part of his life in study and experiment 
without feeling any stress of pecuniary embarrass- 
ment or want, for himself and his family, of the simple 
requirements of existence. In spite of the tardy 
realization of his incessant activity — of the small 
price which he obtained for his masterpieces — ^he was 
passing rich — " rich as Stradivari." 

But times change. Conditions towards the middle 
of the eighteenth century were widely different from 



Conclusion. 13& 



the good old times; artists and artificers could no 
longer aiFord to wait for many months before ex- 
changing work for current coin. A very slow-drying 
varnish must have been a serious obstacle to men of 
small capital, living from hand to mouth ; this con- 
sideration must have driven the violin-makers to seek 
some means of expediting the varnishing operation. 
The deeper-coloured instruments were the first result; 
for not only are the red and red-brown varnishes a 
little harder than those which are yellow and orange, 
but they are also more viscous. The dichroic effects 
become apparent only when the pellicle has reached 
a certain thickness ; with a pale, limpid varnish ten 
coats will be required to be effective, whereas with a 
dark thick one an equally good result may be reached 
with six ; this represents not only a saving of labour, 
but a considerable saving of time, since a week for 
drying must be allowed between each coat. Then, 
again, heating over a water-bath for six or eight 
hours is a small matter with gas at disposal ; with a 
charcoal stove it is a different matter. 

My taste and opinion on the question of medium 
versus dark-coloured varnish, agree with those of 
M. Antoine Vidal ; the dark varnish is the beginning 
of decline (" c'est le commencement de la d6ca- 
dence"). 

There are many excuses, if not reasons, for the 
movement in a downward direction which Stradivari 
himself commenced, in the later years of his life, and 
which some of his pupils soon began to push too far. 
Deep red-brown varnishes were produced which re- 
quired a larger proportion of linseed-oil for solution, 
even then were? very viscous. With time, it appears 



140 Varnishes of the Italian Violin-makers. 

that some of these varnishes have fissured or crackelled, 
while those of Stradivari remain perfect. 

The final evolution of decline is seen in the var- 
nishes of Naples (again a purely arbitrary designa- 
tion). As everyone who knows anything of varnish 
manufacture must be well aware, the preparation of 
the siccative oil is a tedious and troublesome process. 
If the transformation of raw oil is performed by the 
aid of metallic oxides, it requires a long time before 
the boiled oil fi-ees itself, by subsidence, entirely from 
the residue of the oxidant ; it is a matter of months of 
patient waiting. The Neapolitan type appears to owe 
its origin to this inconvenience. 

From my point of view, the Neapolitan varnishes 
were perfectly natural and legitimate developments 
fi-om those of Cremona and Venice, although practical 
difficulties arose which could not, in those early times, 
have been foreseen. 

It occurred to someone that the process of convert- 
ing raw linseed-oil to the siccative form was a similar 
one to that employed for producing resins fi^om the 
turpentines ; therefore, why not combine the two 
operations and thus save time and trouble r 

Colophony was dissolved in raw linseed-oil, or 
Venetian or other turpentine was mixed with it ; this 
solution or mixture was subjected to the same oxida- 
tion and dehydration processes as the turpentines. 

As long as oxidation and dehydration are confined 
within certain limits, the results are perfectly satisfac- 
tory: yellow or orange varnishes prepared in this 
way are transparent, of good colour and of excellent 
quality. But when linseed-oil and turpentine are 
treated together, and the two processes of degradation 



Conclusion. 141 



are carried further, the changes in both do not proceed 
part passu. Moreover, the oil was originally intro- 
duced into the varnish for the purpose of adding to its 
cohesive quality ; if the oil be degraded beyond a 
certain point, it also becomes a resin — the attribute 
which was of so great value to the varnish is lost. 

In making the Neapolitan varnishes, therefore, it 
was necessary to bear in , mind that the linseed-oil 
played a dual part : oxidized up to a certain point, it 
maintained its valuable properties as a siccative oil, 
the yellow colour which it acquired added a little more 
of this colour to that of the resin which made up the 
complement of the pellicle ; oxidized and dehydrated 
beyond this point, the oil became resinous and viscid, 
excellent for aiding the production of varnishes of 
very showy red shades, but no longer serving its 
original purpose. 

In careful hands, under the guidance of an in- 
formed mind, the Neapolitan varnishes give satis- 
factory results; but if these varnishes are made 
without care or knowledge of the principles involved, 
they lack the true properties of an oil- varnish. 

Linseed-oil, by considerable degradation, tends to 
become a substantive red colouring matter, which veils 
(more or less destroys) the refractive eflFects of the 
resin. Consequently, except in yellow varnishes, the 
Neapolitan varieties cannot compare in beauty with 
those of Cremona and Venice. The difference is 
especially apparent on the tables of instruments, the 
delicate grain of which is often obliterated almost 
entirely. 

♦ ♦*«.« 

The origin, gradual improvement, zenith of perfec- 



142 Varnishes of the Italian Vtolin-makers. 

tion, eventual decline of the old varnishes have now 
been traced ; their evolution, deduced as fairly as 
possible from existing records, has been compared 
with that of similar varnishes in my hands. It must 
surely be admitted that if the varnishes, ancient and 
modem, are not the same, I have been misled by a 
most extraordinary chain of coincidences. 

There are yet other idiosyncrasies to be observed 
in both varnishes which are easy of explanation on 
the assumption that they are identical, but otherwise 
difficult to imderstand. It will be noticed, in the 
descriptions of the old varnishes, that, when they are 
pale or yellow, they are generally described as lightly 
and evenly laid on; while the darker varnishes are 
frequently represented as abundantly used, often 
clotted and crusted : their application is supposed to 
have been wanting in care, neatness or skill. I can 
affirm, from my own observation, that this criticism 
(as to the fact) is just. The explanation is simple : — 
The pale varnishes are oxides of turpentine, little 
dehydrated, very soluble in oil of turpentine, limpid, 
easy of even application ; the pellicles are, or appear 
to be, comparatively thin. The darker and brown 
varnishes contain resins much dehydrated, little 
soluble in oil of turpentine; their viscosity, much 
grreater than that of pale varnish, cannot be decreased 
without a large increase of oil. The old masters 
preferred to limit the oil and to do their best with a 
thick, viscous varnish : the results are in evidence. 

Let the critic who is now disposed to cast stones 
at the old artisans come to me ; I will prepare for 
him a fine, viscous, absolutely clear varnish, of either 
type, which, if he essays to lay it on an instrument 



Conclusion. 143 



with perfection of technique, will probably conduce to 
a more charitable frame of mind. My own charity 
has become unbounded. 

♦ « « ' * • 

There is a consensus of opinion that the old 
varnishes were of a soft quality ; the instruments in 
existence appear to afford conclusive evidence of this 
fact, which two centuries have not obliterated or 
concealed. 

All the ancient instruments have the varnish more 
or less removed from parts of the surface of the back. 
This removal of varnish has evidently arisen from 
two principal causes — firstly, the contact of the lining 
of the cases in which they have been preserved or 
conveyed; secondly, the friction of the hands and 
clothing of the players. The " Inlaid Viola " and the 
"Tuscan Violin" exhibit mainly the effects of the 
one cause ; the majority of instruments have suffered 
from both. 

Many people prefer an instrument from which 
parts of the varnish have been so removed — it is sup- 
posed to add to the general picturesque effect. This 
is a matter of taste about which it is proverbially 
useless to argue. 

If one of my instruments (which has been varnished 
for two or three years, carefully finished and polished 
so that the surface of the back is smooth and even as 
that of a mirror) be placed in a closely-fitting case 
lined with silk plush, it may lie for a week or a 
fortnight without any apparent change. After that 
time (the higher the temperature the sooner) a dull 
spot will gradually appear at the point of principal 
contact between the varnish of the back and the 



144 Varnishes of the Italian Violin-makers. 

lining of the case. A few weeks later, the varnish 
will be found to have there received the impression of 
every fibre and inequality of the lining material.- 
Nevertheless — the varnish is perfectly dry and solid — 
it may be repolished. If scraped with a knife, it may 
be removed in fine powder, or it may be chipped off 
in small fragments. 

The workman who fits up the instruments when 
they are well dried often leaves on the varnish — where 
he has unavoidably exerted pressure for a consider- 
able time — ^the exact impression of the skin of his 
thumb or fingers. I have seen on one of the finest 
violins of Stradivari a similar impression of the skin 
of a thumb — ^whether that of the old master or of a 
more modem workman cannot now be conjectured. 

That this impressibility of the varnish is, fi*om 
some points of view, a serious practical inconvenience 
cannot be denied ; no doubt it slowly decreases, 
especially if temperature is maintained above 25° C 
(77° F.). It must be conceded, however, that it is 
impossible to find a varnish permanently elastic in- 
ternally, yet externally capable of resisting continued 
pressure. 

There can be little doubt that this inconvenience 
was felt when saving of time became an object, and 
that it was a factor which affected evolution. That 
Stradivari was well aware of the beneficial effect of 
warmth and aeration is evident fi^om the construction 
of the loft in which he hung his instruments (Hill, 9, 
12). It may be supposed that the temperature in- 
such a loft in the middle of a summer day would 
often exceed 32° to 33° C. (90° F.), with a firee current 



Conclusion. 146 



of air : conditions well suited to the maturation of 
varnish. 

It will be observed, however, that the instruments 
were in the shade, that, apparently, two sides only of 
the loft were open. The notion that the Italians ex- 
posed their instruments to the direct rays of the sun 
is manifestly erroneous ; such a course would be fatal 
to both instrument and varnish. Experts know well 
enough that instruments exist of which the varnish, 
while still immature, has been exposed to the heat of a 
man's body which has been sufficient to cause fissures 
and cracks ; the temperature of the human body never 
exceeds about 38° C. (100° F.), whereas the direct 
rays of the sun, in Italy, may be expected often to 
reach 55° to 60° C. (130*" to 140° F.>. 

It would be interesting to know at what date the 
idea of keeping instruments of this kind in wooden 
cases originated. The oldest lining which I have 
seen is the rough, woollen material known as baize ; 
this material must have affected the soft varnish in 
contact with it in a comparatively short time. The 
modem linings of smooth cloth or of silk plush are 
undoubtedly a great improvement. 

While the modem well-made and sumptuously 
lined cases are admirably adapted for the conveyance 
of instruments, and while, as has been shown, they do 
not suffer in any way from being placed in them for 
a few days, it is open to question whether these cases 
are suited to serve as permanent resting places for 
valuable instruments. Not a few fine specimens have 
received (to my knowledge and regret are still re- 
ceiving) irretrievable damage from the damp and 

V. L 



146 VamtsAes of the Italian Violin-makers. 

mildew which such close cases are liable to engender, 
if instruments are not used for some years. 

The scroll of instruments appears to have been 
designed with a view to afford facility for hanging. 
Formerly it seems to have been customary to hang 
them by the scroll in a suitable cabinet or cupboard, 
or in an apartment, air having free access to every 
part of them. I have found it convenient to revert to 
this ancient usage. 

The introduction of the chin-rest now prevents the 
friction of the chin or beard of the player ; it may also 
be observed that careful persons take means to pre- 
vent clothing or hard substances (such as buttons) 
from damaging their instruments. 

The problem of improvement in the permanent 
resting places of our valued instruments is worthy of 
further consideration, in the direction that has been 
indicated. 

* * « « * 

The conditions attending the preparation of the 
varnishes of Cremona and Venice, as well as the 
inconvenience caused by the length of time required 
for their solidification, may reasonably be considered 
sufficient to account for a gradual modification of their 
constitution, but quite insufficient to justify their total 
extinction in about the middle of the eighteenth cen- 
tury, especially as their loss practically put an end to 
the violin-making industry which had existed in 
Cremona for so long a time. 

But if it is supposed that nitric acid was in fact the 
oxidant employed (there are many considerations 
which support that hypothesis), then a possible 
explanation can be given. 



Conclusion. 147 



It is impossible to avoid the suspicion that, from 
first to last, the Church of Rome not only patronised, 
but assisted and advised the violin-makers by the 
intervention of members of her monastic institutions, 
as she did the architects, painters, and sculptors of 
the period. Gasparo came from Salo, which appears 
to have been, although a small town, a seat of eccle- 
siastical learning where music was cultivated. There 
is a story recorded that Maggini when in dire neces- 
sity was succoured by two monks. It was natural 
and justifiable that the ecclesiastics should patronise 
and encourage, by all means in their power, the arts 
which assisted in the embellishment of their buildings 
and of their religious services. 

It is less easy to understand or to defend the 
interest which they took in the perfecting and promo- 
tion of discoveries and inventions which more properly 
belong to " the world, the flesh and the devil." 

It is asserted that an explosive compound which 
contained nitre was known to the Chinese even before 
the Christian era. In India, some such explosive was 
certainly known long before gunpowder was intro- 
duced into Europe. 

The explosive which Friar Bacon discovered inde- 
pendently, or of which he obtained information fi*om 
his missionary brethren in 1216, was little suited to 
firearms; it was not until the year 1320, when 
Bartholdus Schwartz introduced the process of granu- 
lation, that the history of modem firearms may be 
said to have begun. 

The principal, most costly, ingredient of gun- 
powder is the nitre, a substance which was far fi-om 
abundant in the sixteenth, seventeenth, and eighteenth 

l2 



148 Varnishes of the Italian Violin-makers. 

centuries. The natural stores of nitrate in India, 
China, South America, &c., to which we are now so 
much indebted, were not then available. Its main 
source of supply was artificial; beds of decaying 
nitrogenous organic matter were made imder such 
conditions that the nitrogen was slowly oxidized in 
the presence of available alkali. Every two or three 
years, potassic nitrate was obtained from these beds 
by lixiviation, mixture with wood ashes, subsequent 
purification by crystallization. 

In the manufacture of gunpowder, the almost 
perfect purity of the nitre is a matter of the first 
importance ; its purification from other salts is not 
particularly easy. This subject must have required a 
good deal of patient study. The improvement of 
gunpowder must have entailed the attainment of very 
considerable knowledge of potassic nitrate and its 
reactions with other chemicals, such as sulphur 
(another of the constituents of gunpowder). 

Cremona formerly belonged to the Duchy of Milan. 
The last of the Sforza, Dukes of Milan, died in 1535. 
The Duchy then passed under the rule of Spain, and 
appears to have enjoyed comparative peace until the 
beginning of the eighteenth century (fighting in a 
small way, pestilence and famine were ordinary occur- 
rences). The great War of the Spanish Succession 
then began, involving several of the European Powers 
in a long and sanguinary struggle. Cremona itself 
did not escape the horrors of war. By the treaty of 
Utrecht in 1 7 1 1 and 1 7 1 3 the Duchy of Milan was 
ceded to Austria. Shortly before the middle of the 
eighteenth century another great European War broke 
out^-that of the Austrian Succession. 



Conclusion. 149 



The demand for nitre in Europe must have largely 
increased during the wars of the eighteenth century. 
The efforts made by Governments to assure a supply 
of nitre for the manufacture of explosives were such 
as can hardly be conceived possible at the present 
time. It was no unusual thing for a European State 
to enact that the peasant-farmers should pay a defined 
proportion of their taxes in saltpetre or nitre, which 
they were obliged to obtain by the artificial process 
already described ; such enactments have remained in 
force in some countries during my lifetime. 

What means were taken by Austria, within her 
dominions, for the necessary supply of nitre to the 
Government during the War of Succession cannot 
now be easily ascertained (her geographical position 
in relation to the question must not be ignored) ; it is^ 
at least probable that decrees may have been promul- 
gated and enforced without any regard to the influence 
of such proceedings on commerce or the industrial 
arts. 

Naples and other towns (such as Bologna) which 
were not under Austrian rule may have been in a 
totally different position as to the supply of nitre, and 
consequently of nitric acid. 

It is significant that gunpowder appears to have 
been, up to the middle of the nineteenth century 
(perhaps later), one of the State monopolies of 
Austria. 

Whether the occurrence of the War of Succession 
and the sudden extinction of the Cremonese varnish 
at the same period was a mere coincidence, or whether 
the one had an influence on the other, must be left to 
the reader to judge firom the brief facts which have 



160 Varnishes of the Italian Violin-makers. 

been stated ; the question may perhaps be considered 
worthy of further research. 

* • ♦ ♦ ♦ 

The question as to whether there existed a secret 
or not, is not material ; but it is of interest, because a 
secret may have had some influence on the sudden 
extinction of the varnish, and may account, under 
certain circumstances, for the complete loss of its 
method of preparation. 

M. Mailand and others have asserted, without 
much reason, that the maintenance of such a secret, 
for so long a time, would be impossible ; the assump- 
tion that similar varnish, or varnish formed from the 
same basis, was in common use or was to be found in 
commerce, has been shown to be entirely without 
foundation. 

If the process of manufacture were that which I 
have adopted, there would be no difficulty in keeping 
the secret ; I could easily have kept it to myself if I 
had been so disposed. If I divulged it to others, why 
should they make it public if their interest and liveli- 
hood depended on the maintenance of secrecy ? The 
varnish itself, whether in its fluid state or in the 
pellicle, affords no clue whatever as to its origin. 
Even the most accomplished chemist of to-day would 
be unable to discover how it was produced, much 
less one of the sixteenth, seventeenth, or eighteenth 
century. 

There happens to be an analogous case of a secret 
manufacture, which has been well preserved in these 
later days, connected with the manufacture of glass 
for optical purposes. The extraordinary thing is that 
this highly technical discovery should have been 



Conclusion. 161 



made by a simple, but doubtless very intelligent, 
workman, and that men of science are still unable to 
ascertain the exact detail on which success depends. 
This discovery was made by P. Guinand (the son of a 
working carpenter, of Brenets, in Canton Neuchatel, 
Switzerland), who was bom in 1740. The secret was 
communicated to several persons, but, nevertheless, it 
has not leaked out. (Cantor Lecture : " Glass for 
Optical Purposes" by Richard T. Glazebrook, M.A., 
D.Sc, F.R.S. Lecture I. Journal of the Society of 
Arts. 17th October, 1902.) 

What evidence is there to show that the Italian 
violin-makers were not members of a " Guild " ? It 
may have been a secret society ; that would be well in 
accordance with Italian methods. The use of labels 
from the earliest times is consistent with its existence, 
so are the relations which appear to have existed 
between master, pupil or apprentice and journeyman. 
Trade Guilds were much in vogue at the period. 
M. Vidal g^ves an interesting account of several of 
these Guilds (in other countries), with their rules or 
statutes. Only approved artificers were admitted to 
membership ; the laws to which they had to submit 
were strict, sometimes arbitrary. The rules of such a 
Guild may explain the reason why Antonio Stradivari, 
for so many years of his life, used the labels of Nicol6 
Amati. Only members of the Guild were allowed to 
use such labels ; instruments without the label of a 
member were not permitted to be sold, under severe 
penalties. 

The existence of a Guild secret woidd enable a 
reasonable explanation of the extinction of the var- 
nish and process to be gfiven. If the varnish secret 



162 Varnishes of the Italian Violin-makers. 

were known only to a certain number of persons, who 
were not initiated until they became "masters" of 
the guild — t,e.j until they had reached full manhood — 
and if the use of the secret were arrested for a con- 
siderable period from the scarcity of an essential 
ingredient, it is quite conceivable that the process 
might die out entirely ; those who had practised the 
art would be driven, by necessity, to seek other var- 
nishes, other methods; there would no longer be 
any reason (or any means) for teaching the rising 
generation ; the cognoscenti would gradually die out, 
canying the knowledge of an obsolete process with 
them to their graves. 

M. Alphonse Daudet ("Tartarin de Tzu-ascon"), 
who ought to be a competent authority, expresses 
the opinion that people of the North are not qualified 
to pronounce judgment on evidence emanating from 
the South ; a person having very little acquaintance 
with the sunny South must not venture therefore to 
express an opinion on the stories relating to this 
matter which havQ been circulated; but he may be 
allowed to state that he is in the habit of basing his 
opinions on evidence of a different character. 

« « « « « 

Another point requires some consideration. It 
has been constantly asserted that the varnishes in 
question were in common- use for other purposes at 
the period when they were employed by the violin- 
makers. M. Mailand, with easy access to Continental 
museums and collections of antique furniture and 
cabinet-work, was perfectly well aware that examples 
of this kind of workmanship, covered with the varnish, 
were not in existence ; he theiiefore had to seek some 



Conclusion. 163 



other support for this theory. He admits that the 
varnishes which the violin-makers used, "whatever 
" they might be, were excellent for this special appli- 
" cation, since the instruments which they covered 
** with them have reached us after more than two 
" hundred years in a fine state of preservation, and 
" since these varnishes have protected them, while 
" leaving them an entire liberty of vibration " (9, 10). 
. . . . But, "In effect, at the period in which 
" the Italian masters worked, the fabrication of var- 
" nishes was still in its infancy ; however, these men, 
" who knew their art so well, recognized that those 
" which were made in their time and which they had 
appropriated to their use, although bad for other 
purposes (quoique mauvais pour d'autres besoins), 
were precisely those which best suited their in- 
struments" (4). He thinks that French polish, 
varnishes made fi-om copal, amber, and similar sub- 
stances, because of their greater solidity, were vastly 
superior for all other purposes, but not for instru- 
ments. 

This argument (founded on a fair amount of 
imagination) has a hole in it! If the varnish was 
common, then, besides violins, other wooden articles 
covered with it should and must have reached us ; if 
other articles so covered have not reached us, then, 
in order to support the assumption of opportunity for 
general use, it must be clearly shown that the other 
artificers had other varnish better suited to their 
purpose. 

Copal- and amber -varnish, French polish, and 
spirit varnish in general are beside the question ; for 
it is quite certain that the cabinet-makers of the 



a 
a 



164 Varnishes of the Italian Violin-makers. 

sixteenth century did not either prefer or reject 
them^ for the simple reason that they were not in 
existence at that time. M. Mailand proves to his own 
satisfaction that the experts of Bonanni's time (17 13) 
coiild not dissolve either hard copal or amber ; lac he 
states to have been first introduced into Europe 
shortly after the middle of the seventeenth century ; it 
was probably many years later before it was com- 
monly used in alcoholic solution. It will be remem- 
bered that^ according to M. Vidal, Louis Guersan (a 
pupil of Jacques Boquay) was the first French violin- 
maker to try spirit varnish for his instruments, some 
time after 1735. Did the cabinet-makers of the 
sixteenth century, in fact, use varnish which was 
more solid than that of Cremona ? M. Mailand him- 
self gives information as to the kinds of varnish 
commonly known at that time, by quoting fi"om the 
publications of Alexis (1550) and Fioravanti (1564). 
The former prescribes potable spirit [eau de vie) for 
the solution of benzoin ; the latter, the same men- 
struum for benzoin, sandarac, and mastic ; this solvent 
was clearly available only in small quantity, and is 
not strong enough for an effective spirit varnish. The 
ingredients of the oil-varnishes were mastic, sandarac, 
pine-resins of different kinds, linseed-oil and turpentine 
(the three latter were the preponderating constituents). 
It is certain that, fi"om these, oil-varnish harder than 
the Cremonese could not be obtained ; quite the reverse. 
A very little study of the prescriptions given by the 
various authors cited is sufl5.cient to convince anyone 
that the best varnish capable of production by the 
directions given would be vastly inferior in every 
respect to the varnish used by Maggini, the Amati, 



i 



Conclusion. 165 



the Guamerii and by Antonio Stradivari. The reason 
why the cabinet-makers did not use this varnish 
instead of tlie very inferior ones at their disposal was 
simply because they were unable to obtain it. 

Now as to the main assertion that Cremonese and 
Venetian varnishes were and are unsuitable for other 
purposes. Prima faciCy it must be evident that the 
varnish on a violin or violoncello in daily use, which 
M. Mailand acknowledges as efficient for the preserva- 
tion of these instruments for two centuries, is solid 
enough for the protection and fine enough for the 
embellishment of the cabinet in which the instruments 
hang, and for other similar articles of furniture which 
are not exposed to more severe friction and wear than 
the instruments. 

Prior to the introduction of French polish, chairs 
and tables were probably polished with the simple 
compound of beeswax, linseed- oil and turpentine, 
which was generally used for that purpose in Europe 
up to about the beginning of the last century ; it is 
readily admitted that, except for the top of a dining- 
table and for other articles of furniture similarly 
exposed to heat, French polish is an excellent sub- 
stitute for the crude, simple and laborious polish of 
the olden time. 

The only way to put an end to these assertions, 
assumptions and theories as to the unsuitability of 
Cremonese varnishes for ..purj^oses other than the 
covering of instruments, was to make a practical 
experiment on a sufficiently large scale. A spacious 
music-room was built of stone; it was ceiled with 
wood decorated with wooden mouldings in geometric 

9 

patterns ; the walls were lined and panelled with 



166 Varnishes of the Italian Violin-makers* 

selected wood of different varieties. The choicest 
copal varnishes that could be obtained, of various 
shades, were used to cover parts of the dressed wood ; 
other parts were French polished ; the remainder (as 
well as cabinets for instruments, bookcases and shelves 
for printed music) was covered with Cremonese var- 
nish of my own preparation. The different varnishes 
can now be seen and examined, side by side, imder 
identical conditions. The result is conclusive that 
the Cremonese varnishes are perfectly suitable for the 
purpose ; as may be supposed, their decorative effect 
is imapproachable. 

As to hardness, that is a question of extent of 
degradation, and of the kind of turpentine chosen as 
raw material ; Venice turpentine producing, other 
conditions being equal, more solid varnish than pine- 
turpentine. The reason for this difference has yet to 
be ascertained. As to durability, for interior use the 
amount of oil used for instruments is suflB.cient ; for 
exterior use, my short experience and a few experi- 
ments lead me to suppose that increase of oil is 
necessary to enable any varnish to withstand weather 
and climate. The ing^redient which resists these in- 
fluences being the oil, and not the resin, of whatsoever 
kind it may be, the condition of the oil (as to extent 
of degradation) is a factor which must not be ignored. 
* « « » « 

The imagination contemplates with pleasure gene- 
rations of artizans, leading simple and industrious 
lives, patiently seeking to bring to perfection the 
various instruments of the violin family which were 
destined to give unalloyed enjoyment to succeeding 
thousands of men and women. Malignant rumour 



Conclusion* 157 



has indicated the existence of one black sheep among 
their number ; but, happily, the vague stories which 
have been promulgated are apparently apocryphal, 
to say the least, they are of extremely doubtful 
authenticity. Antonio Stradivari, above all, has left 
behind him a record of a man of eminent ability and 
intelligence, of steadfast purpose and untiring in- 
dustry, pursuing a lofty ideal during a long, unosten- 
tatious and blameless life. 

Suddenly this charming reminiscence is lost in 
oblivion. In a few years humanity began to realise 
that they had sustained a loss which appeared 
irreparable. A struggle began for the possession of 
the instruments which the old masters had be- 
queathed to posterity. The old condition of industry 
and simplicity was changed to one of craft and 
deception. 

M. Vidal gives a vigorous description of the state 
of things which ensued. The most prominent and 
gfrotesque actor in the scene is Luigi Tarisio, an 
Italian pedlar of low origin, who, apart from cupidity, 
appears to have had a passion for the acquisition of 
the old instruments, the best of which he could not 
bring himself to part with. At his death, in his 
miserable resorts, he left behind him a hoard of about 
250 instruments, many of them perfect masterpieces, 
which realised an unexpected fortune for his poverty- 
stricken heirs. In M. Vidal's book will be found a 
graphic and amusing description of the man and his 
methods. 

The demand caused by the eagerness of artistes 
and amateurs for specimens by masters of the greatest 
renown soon led to its supply. Labels were exchanged 



158 Varnishes of the Italian Violin-makers. 

and forgedj'instruments were made up of heterogeneous 
parts. For many years the best workmen of Europe 
devoted their energies and talents to the task of pro- 
ducing counterfeits of the old masterpieces, sacrificing 
their artistic individuality to a degrading traJBELc in 
order to obtain the necessaries of modem existence. 
While every charitable mind will seek to find excuses 
for these proceedings, it is futile to attempt to dis- 
guise the fact that the originating cause was that they 
found themselves able to obtain for copies (including 
facsimiles of the labels and imitations of the effects of 
age and wear) a higher price than their own acknow- 
ledged work would bring. If they were not actual 
perpetrators of firaud, they brought the possibilities of 
it within the reach of ignorant or unscrupulous 
dealers. The young artist who copies pictures by 
old masters for the purpose of increasing his own 
knowledge and skill, signing the copies with his own 
name, does that which is perfectly legitimate and 
proper; but an experienced and accomplished artist 
who makes a constant practice of copying ancient 
masterpieces, imitating the signatures and the ap- 
parent effects of time and accident, does that which 
may be excused by extenuating circumstances, but 
cannot be defended by any sophistry. 

Under the sinister influences described, there were 
both violin-makers and dealers that stood firm to 
settled principles of probity and ingenuousness ; but, 
unfortunately, there is too much reason to suppose 
that they were occasionally duped by the productions 
of their more astute and less scrupulous competitors. 
The confusion and distrust arising fi-om this state of 
things was bad enough ; what it might have been had 



Conclusion. 159 



the concocters of counterfeits been in a position to use 
the proper kind of varnish can well be imagined ; the 
varnish was the weak part of the illicit productions 
which betrayed them. 

With the vast increase of wealth during the latter 
half of the nineteenth century, as well as the growing 
numbers of artistes and amateurs, the value of good 
examples of the work of the old Italian masters has 
been steadily enhanced; it is not surprising that 
trafl&c in instruments has proved to be a more lucra- 
tive employment than the production of genuine 
modem ones: the violin-maker, as an artist, has 
practically disappeared. 

There are, nevertheless, instruments made to-day 
which, as far as wood and workmanship are con- 
cerned, are admirable in every respect. The restora- 
tion to the artistic violin-maker of the varnish which 
has been so long withheld from him — ^the increment 
which was wanting to enable him to attain his ideal 
of perfection — is a prospect which cannot be contem- 
plated without satisfaction. 

If the views which have been expressed (which 
originated in the researches of M. F61ix Savart) as to 
the causes of the deterioration of the sounding-boards 
of pianofortes are well founded, it must be evident 
that the spirit varnish at present used for them is 
not the most suitable; it may be expected from all 
the evidence that the substitution of Cremonese or 
Venetian oil-varnishes for this purpose must result 
in improvement of tone when these instruments are 
new, and in the prevention of deterioration as they 
grow old. 

With increasing education and enlightenment. 



160 Varnishes of the Italian Violin-makers. 

civilized people will more and more perceive that 
artistic productions which are base in design or in 
execution do not necessarily become more desirable 
because they are old. The value of a masterpiece 
becomes enhanced by age because of its inherent 
excellence and its rarity; mean productions become 
neither admirable nor rare with increasing age. 

The dogma that instruments of the violin family 
improve by age has been so persistently promulgated 
that it has received general credence; my efforts to 
obtain real evidence in its support have met with an 
entirely negative result : whatever evidence exists on 
the question of tone points in the opposite direction. 

Experiments on this subject hav^, in the opinion of 
M. Vidal, given rise to the most remarkable pheno- 
mena in the history of bow instruments. He relates 
at length what occurred (68, 69, 70, 71). M, Fran9ois 
Chanot (" Capitaine ing^nieur de deuxi^me classe dans 
" la marine fran9aise ") brought before the Acad6mie 
des Beaux- Arts of Paris, in 181 7, a violin of a new 
form which he had invented. The Acad^mie appointed 
a commission to judge of its quality of tone. The 
members of this commission were de Gossec, Cheru- 
bini, Catel, Lesueur, Ch. de Prony and Berton. M. 
Alex. Boucher was chosen as the player, and, in order 
that there should be no bias, he played on a fine violin 
of Stradivari and on the new violin in an apartment 
adjoining that wherein the judges were assembled, so 
that he could be heard without being seen. "The 
** whole commission, in three consecutive trials, always 
" thought they heard the Stradivari when M. Boucher 
•* was playing on the new violin, and vice versa when 
" he played on the Stradivari." The question was 



Conclusion. 161 



decided in favour of the violin of M. Chanot, which, 
although new and made of fresh wood, was able to 
bear successfully this trying ordeal. The Academie 
seem to have been so much surprised at this result 
that they demanded a new trial before their members 
(in general meeting) on the 26 July, 181 7. Boucher 
repeated his performance with the same result — ^the 
Chanot violin was adjudged to be much superior to 
the Stradivari. 

M. F61ix Savart obtained an identical verdict, in 
1 81 8, for a violin of his invention. The Commission 
on that occasion consisted of Ch. de Prony, Cherubini, 
Catel, Berton, Lesueur, and Biot ; the player was 
M. Lefebvre. 

Nevertheless the prestige of these two new violins 
lasted only two or three years. 

The conclusions to be drawn from these experi- 
ments, which appear to have been conducted under 
such conditions as to eliminate all prejudice or bias, 
are either (i) that the judges were deceived and came 
to a wrong conclusion, which seems scarcely credible; 
or (2) that the Stradivari had not improved by age, 
and (3) that the new violins deteriorated in a very 
few years. 

Similar experiments have been made by myself 
under the same conditions. The result has been to 
prove that musicians with a cultivated ear cannot 
distinguish between the tone of a new instrument and 
that of those which have been in existence for about 
two centuries. 

All the available evidence is in favour of the sup- 
position that violins do not improve with age. My 
firm conviction is that instruments which are at 

V. M 



162 Varnishes of the Italian Violin-makers. 

present mediocre or inferior, either in appearance or 
tone, so far from becoming better in either respect 
with use and age, will decidedly deteriorate, even if 
they exist for a century or more. Whether instru- 
ments which are now good in both respects will 
maintain these qualities for a great number of years 
must remain an open .question. 

It would be some satisfaction to know that in the 
near future an excellent instrument will be within the 
reach of any one who seriously desires it and has the 
intelligence to appreciate it, without troubling our 
minds about its transmission to posterity. The first 
thing we have to do is to produce instruments equal 
to the old ones. If we succeed in doing this we may 
reasonably expect that the modem ones," like the 
others, will bear the test of time. What has been 
done once may be done again. To be satisfied with 
productions which are admittedly inferior to the old 
ones, flattering ourselves with the hope that the 
beneficent hand of time will improve them, is to set 
up for ourselves the pursuit of a m3rth. This is cer- 
tainly no Stradivarian method. He would not have 
bequeathed to us either unsurpassed masterpieces, or 
the still more valuable example of a life of lofty and 
steadfast purpose, if he had contented himself with 
the indulgence of a vague and fantastic chimera. His 
own works show that his method was based on crucial 
tests of his own productions, which led to gradual 
progress towards his ideal of perfection. 

General commiseration is felt for the unfortunate 
artiste who, of necessity, attempts to convey the 
expression of thought and - sentiment to a critical 
audience through the medium of an indifferent instru- 



Conclusion. 163 



ment. My deepest sympathy attends him in his 
hours of study and practice. He must be a genius 
indeed who, under such adverse conditions, distances 
his rival who has the inestimable educational ad- 
vantage arising from the possession of a sympathetic 
and resonant violin or violoncello. In my humble 
opinion, a fine instrument is fiiUy as important a 
factor in education as it is in display. 

4F « » » « 

The processes herein described are the subjects of 
Letters Patent in Great Britain and Ireland No. 
19,626 of 1902 ; and in the United States of America 
No. 754,298 of 1904. 



m2 



( 164 ) 



BIBLIOGRAPHY. 



Berthslot. Annales de Chimie. Vok. 38, 39, 40. Paris, 1853, 1854. 

DiETESiCH, Dr. Karl. Analysis of Resins (trans, by Chas. Salter). 

London. 1901. 

Engbl, Carl. Researches into the early history of the Violin family. 

London, 1883. 

F^Tis, F. J. Biographie universelle des musiciens. Paris, 1889. 

Fleming, James M. Old violins and their makers. London, n.d. 

♦Gallay, J. Les luthiers italiens (nouvelle Edition du 

Parfait Luthier de Pabbe Sibire). Paris. 1869. 

Groye, Sir George. Dictionary of music and musicians 

(a.d. 1450-1889). London. 1900. 

*Ha&t, George. The violin, its famous makers and their imitators. 

London. i88o. 

Hawsis, Rev. H. R. Old violins. London. 1898. 

^Helmholz, Hermann L. F. On the sensations of tone 

(trans, by Alex. J. Ellis). 2nd edit. London, 1885. 

*HiLL, Messrs. A short account of a violin by Stradivari, dated 1690. 

London. 1889. 

* do. do. Gio : Paolo Maggini, His Ufe and work. London. 1892. 

* do. do. Antonio Stradivari, His life and work (1644-1737). 

London, 1902. 

HiPEiNS, A. J. Musical instruments, historic, rare and imique. 

Edinburgh. 1888. 

*Mailand, Eugene. D6couverte des anciens vemis italiens 

(nouvelle Edition). Paris. 1874. 

*Naudin, Laurent. Fabrication des vemis. (Gauthier-Villars.) 

Paris. 1902. 

Naxtmann, Emil. History of music (trans, by F. Praeger). 

London, N.D. 

Otto, Jacob A. A treatise on the structure and preservation of the viohn 

(trans, by John Bishop). London. 1875. 

Rabat^, E. L'industrie des resines. (Gauthier-Villars.) 

Paris. 1902. 

TiNGEY, P. F. The painter and vamisher's guide. (Second edition.) 

London. 1816. 
*ViDAL, Antoine. Les instruments ^ archet. Paris. 1877. 



( 165 ) 



APPENDIX. 



DETAILS OF THE COMPOSITION OF VARNISHES 
WHICH HAVE BEEN MADE AND USED FOR 
COVERING MUSICAL INSTRUMENTS AND FOR 
OTHER PURPOSES. 

Colophony and turpentine of Venice are measured by weight (grammes) ; 
Nitric acid, linseed-oU, oil of turpentine by volume (cubic centimetres). 
Turpentine of Venice is estimated at 80 per cent, as resin. 
Oil of turpentine is estimated at 50 per cent, as resin. 

CREMONESE VARNISH. 

No.l. 

Made 13 February^ 1901 {dry way). 

Farts. Per cent. 

Colophony loo 28*57 

Nitric acid 20 

(oxidation reaction) 

Linseed-oil (boiled) . . . . 50 • 14*29 

(heated over sand-bath to 120^ C.) 

Oil of turpentine 200 — 240 57*14 

* 350 100*00 



No. 2. 

Made 19 February^ 1901 {wet way ^ sand-bath ^ but not 

above 100° C). 

Parts. Per cent. 

Colophony 100 28*17 

Oil of turpentine .... 50= 25 7*05 

Nitric acid 20 

(oxidation reaction) 

Linseed-oil (boiled) . . . - 50 14*08 

Oil of turpentine 180 — 240 50*70 

355 100*00 



166 Appendix. 

CREMONESE VARNISH. 

No. 3. 

Made 17 Marchj 1901 {^ater-bath and water -oven). 

Parts. Per cent. 

Colophony 100 23-81 

Oil of tuipentine .... 60= 30 7*14 

Nitric acid 20 

(oxidation reaction) 

Linseed-oil (boiled) . . . • 50 ii'qi 

(heated in water-oven (lOO^ C.) for 5 hours) 

Oil of turpentine 240 57-14 

420 100*00 



No. 4. 

Made 19 Aprils 1901 {sand-bath). 

Parts. Per cent. 

Colophony 100 27-78 

Oil of turpentine .... 60= 30 8-33 

Nitric acid 20 

(oxidation reaction) 

Linseed-oil (boiled) .... 50 ^y^9 

(maintained at a temperature of 125 — 130^ C. 
for about 15 minutes) 

Oil of turpentine 180 50-00 

360 lOO'OO 



No. 5. 

Made 27 Aprils 1901 {jvater-bath and water-oven). 

Parts. Per cent. 

Colophony 100 27-78 

Oil of turpentine .... 60= 30 8-33 

Nitric acid . . . . • 3^ 
(oxidation reaction) 

Linseed-oil (boiled) .... 50 13-89 

(heated in water-oven about 6 hours) 

Oil of turpentine 180 50-00 

360 100*00 



Composition of Varnishes. 167 



CREMONESE VARNISH. 
No. 6. 

Made 4 September ^ 1901 {water-bath and water-oven). 

Parts. Per cent. 

Colophony 100 25*64 

Oil of turpentine .... 60= 30 7*69 

Nitric acid . . . . • 5^ 
(oxidation reaction) 

Linseed-oil (boiled) .... 50 12*82 

(heated in water-oven 6 to 7 hours) 

Oil of turpentine 210 53*85 

390 lOO'OO 



No. 7. 

Made 9 December j 1901 (water-bath and water-oven). 

Parts. Per cent. 

Colophony 100 25-97 

Oil of turpentine .... 70= 35 9*09 

Nitric acid 33'3 

(nitrification extended over 4 days) 
(oxidation reaction) 

Linseed-oil (boiled) .... 50 12-99 

(heated in water-oven about I hour) 

Oil of turpentine 200 5^*95 

385 lOO-QO 



No. 8. 

Made 9 December ^ 1901 {second timi). 



Colophony .... 
Oil of turpentine 

Nitric acid .... 
(nitrification extended over 4 days) 
(oxidation reaction) 

Linseed-oil (boiled) 

(heated in water-oven about 6 hours) 

Oil of turpentine 



Parts. 


Per cent. 


. 100 


28-17 


70= 35 


9-86 


IVl 




60 


16*90 


. 160 


45-07 


355 


100-00 



168 Appendix* 



VENETIAN VARNISH. 

No. 9. 

Made 21 J^ebruary, 1901 ipater-bath). 

Parts. Per cent. 

Venice turpentine .... 100= 80 26*67 

Nitric acid 20 

(oxidation reaction) 

Linseed-oil (boiled) . . . .40 ^yiZ 

(heated over water-bath i hour) 

Oil of turpentine 180 60*00 

300 100*00 



No. 10. 

Made 24 Mayy 1901 {pater-iath and water-oven). 

Parts. Per cent. 

Venice turpentine . . . . 100= 80 30*30 

Nitric acid 24 

(oxidation reaction) 

Linseed-oil (boiled) . . . .40 ^S'^S 

(heated in water-oven about 6 hours) 

Oil of turpentine 144 54*55 

264 100*00 



No. 11. 

Made 4 September y 1901 {pater-bath and water-oven). 

Parts. Per cent. 

Venice turpentine .... 100= 80 31*58 

Nitric acid 33*3 

(oxidation reaction) 

Linseed-oil (boiled) .... 33*3 '3*15 

(heated in water-oven 6 to 7 hours) 

Oil of turpentine . ^ . . . 140 55*^7 

253*3 loo-oo 



Composition of Varnishes. 169 



VENETIAN VARNISH (HYBRID). 

No. 12. 

Made 29 November ^ 1901 {pater "bath and water ^oven). 

Parts. Per cent. 

Venice turpentine . . . . 100= 80 27*40 

Oil of turpentine .... 16= 8 274 

Nitric acid 32 

^nitrification extended over 4 days) 
(oxidation reaction) 

Linseed-oil (boiled) .... 44 15*06 

(heated in water-oven about 5 hours) 

Oil of turpentine . . . . . 160 . S4*8o 

292 100*00 



No. 13. 

Made 25 Aprils 1902 {pater-bath and water-oven). 

Parts. Per cent. 

Venice turpentine .... 100= 80 24*84 

Oil of turpentine .... 100= 50 iS*53 

Nitric acid . . . . .64 
^nitrification extended over 3 weeks) 
(oxidation reaction) 

Linseed-oil (boiled) .... 48 14*91 

(heated in water-oven about 5 hours) 

Oil of turpentine 144 4472 

322 100*00 

V. N 



7 



170 Appendix. 



NEAPOLITAN VARNISH. 

No. 14. 

Made lo March^ 1901 i^ater-iath). 

Parts. Per cent. 

Colophony loo 28*17 

Linseed-oil (raw) 25 7*04 

Nitric acid 20 

(ozidation reaction) 

Linseed-oil (boiled) .... 50 14*09 

(heated over water-bath about 7\ hours) 

Oil of turpentine . . . . .180 S070 

355 loo'oo 

No. 15. 

Made 8 Aprily 1901 {sand-bath). 

Parts. Per cent. 

Colophony 100 29*20 

Linseed-oil (raw) 62*5 18*25 

^^ Nitric acid 12*5 

^^^ ^oxidation reaction} 

00 (heated over sand-oath until of a deep brown 

._■ - colour) 

js^ Oil of turpentine 180 52*55 



342*5 loo-oo 

TO 

:|NO No. 16. 

Made 30 August^ 1901 (pater-bath and water-oven). 

Parts. Per cent. 

Venice turpentine .... 100= 80 14*50 

Colophony 80 14*50 

Linseed-oil (raw) 16 2*90 

Oil of turpentine . . . . 48= 24 4*35 

Nitric acid 48 ' 

^xidation reaction) 

(heated in water-oven about 6 hours) 

Linseed-oil (boiled) . . . .64 11 '60 

Oil of turpentine 288 52*15 

PRINTED BY C. F. ROWORTH, GREAT NEW STREET, FETTER LANS, E.C. 



THE BORROWER WILL BE CHARGED 
AN OVERDUE FEE IFTHIS BOOK IS NOT 
RETURNED TO THE UBRARY ON OR 
BEFORE THE LAST DATE STAMPED 
BELOW, NON-RECEIPT OF OVERDUE 
NOTICES DOES NOT EXEMPT THE 
BORROWER FROM OVERDUE FEES. 



I